HPA-axis and insulin/Glucose levels during a course of electroconvulsive therapy

Introduction Electroconvulsive therapy (ECT) is one of the most effective treatment options for therapy resistant psychiatric disorders. Patients suffering from obesity seem to respond well to ECT and even weight loss was reported. In nondiabetic patients, an increase in blood sugar of about 10% as well as an increase of insulin after one ECT treatment was reported. Chronic ECT was shown to be related with a decline in insulin levels in respondersʼ only. The underlying mechanism is unknown and could be related with seizure-induced hormone secretion such as catecholamines or cortisol 4. The aim of our study was to assess cortisol, metanephrine, normetanephrine, glucose and insulin during a course of ECT in remitters and non-remitters.

Methods 12 men and 20 women receiving ECT were included in our study. Outcome analysis was performed for patients with pharmacoresistant major depressive disorder (n = 28). Remission was defined as a MARDS score below 10. Blood was withdrawn directly before and 15 minutes after the first and directly before the last ECT. Glucose levels were assessed using enzymatic reference method with hexokinase. Cortisol and insulin levels were assessed using electrochemiluminescence immunoassay (ECLIA). Metanephrine and normetanephrine were measured using HPLC.

Results At baseline, there were no significant differences between remitters and non-remitters to ECT. In remitters, normetanephrine levels differed significantly compared to non-remitters (p = 0.049). 15 minutes after ECT, glucose, cortisol and normetanephrine increased significantly, when compared to baseline (glucose: p < 0.001; cortisol: p < 0.001 normetanephrine: p = 0.018), but there was no chronic effect for glucose and normetanephrine (glucose: p = 0.902; normetanephrine: p = 0.998) Cortisol showed a significant difference between 15 minutes after the first ECT and before the last ECT (p = 0.002).15 minutes after the first ECT, there was a trend towards higher glucose and insulin levels in remitters (n = 13) to ECT compared to non-remitters (n = 19) (glucose: p = 0.201; insulin p = 0.118). Metanephrine did not differ between remitters and non-remitters (before and after first ECT p = 0.291). Correlation analysis revealed a positive correlation between duration of episode and cortisol level (r = 0.639; p = 0.006). Baseline Beck-Depressions-Inventar-II (BDI-II) was positively correlated with normetanephrine levels (r = 0.389; p = 0.037) and CRP (r = 0.402; p = 0.031) and we found a positive correlation between normetanephrine and baseline heart rate (r = 0.506; p = 0.003).

Conclusion Previous reports showed a reduction in urinary (nor-)metanephrine secretion in ECT patients. Our results show the acute effect in plasma rather than the effect during 24 hrs. The findings concerning glucose and insulin go in line with previous reports, even though we could not find significant differences but trends concerning response and glucose/insulin. It is possible, that the release of insulin reflects neural activation during ECT 4, since the pancreas underlies neural control of the right vagal nerve.