INTRODUCTION
Late life depression is a highly prevalent psychiatric disorder worldwide, that leads to great disability, disease burden, and health costs [1,2]. Considering the increase in the number of elderly people, health and emergency services need to be increasingly prepared to meet their psychological needs [3]. Nonetheless, several treatment modalities, including antidepressants, somatic therapies, psychosocial interventions, neurobiology based psychotherapy, and others, exist for treating late life depression [4]. In the elderly population, the efficacy of antidepressants might decrease [5], and response to electroconvulsive therapy (ECT) might increase [6]. Treatment-resistant depression in late life depression remains a significant public health problem. ECT, with a remission rate of 60%–80%, is of utmost importance for improving the quality of life, alleviate suffering, and prevent suicide in such patients [7]. ECT is arguably the most effective treatment for severe major depression, with response rates and times superior to other available antidepressant therapies. Neuroimaging research suggested that hippocampal plasticity after ECT plays an important role in the antidepressant response to ECT, which is in line with the evidence of upregulated neurotrophic processes in the seizure physiology [8]. ECT is, however, underutilized in the geriatric population [9] owing to multiple reasons, including the disadvantage of disturbance in short-term memory [10]. For maximizing efficacy while minimizing cognitive side effects, modern ECT includes individualized stimulus dosing, electrode placement, pulse amplitude and width, anesthesia regimens, and/or administration frequency. The harmful consequences of ECT, such as post-ECT delirium and anxiety about adverse cognitive effects sometimes leads to its premature discontinuation [11], which necessitated the development of a novel ECT strategy. Clinical observations (as well as animal models) in situations that ECT produced only some or no seizure activity revealed that low energy stimulation may be effective [12-14]. Low energy ECT exhibited fewer adverse cognitive effects but lower efficacy than conventional ECT [12]. Most double-blind clinical trials demonstrated that the greatest therapeutic response of ECT is observed during the first three sessions [15,16]. To combine the advantages of conventional ECT stimulus and low energy ECT, a pilot randomized controlled trial was conducted using mixed-strategy ECT (msECT), in which the first three sessions were completed using conventional bitemporal ECT to maximize antidepressant efficacy and the subsequent sessions using low energy bitemporal ECT to minimize side-effects. In a previous study [14], the dosage in low energy ECT was set to 5% energy with the DGX mode (pulse width=1 ms; frequency=30 Hz), and the results showed a therapeutic response, suggesting an effective alternative electric stimulus-based treatment. Therefore, in the present study, this low energy ECT strategy was adopted for the later stages. The study aimed to evaluate whether low dosage during the subsequent sessions of msECT results in comparable efficacy but fewer side effects than regular ECT (RECT) among patients with late life depression.
METHODS
Overview
This randomized controlled trial with three follow-up visits was conducted from 2018 to 2021, at a single site, Anhui Mental Health Center (The Fourth People’s Hospital in Hefei), in Hefei city, Anhui province, China. Participants provided written informed consent prior to their inclusion in the study. This trial was designed in accordance with the Declaration of Helsinki, approved by the Hefei Fourth People’s Hospital’s medical ethics committee (2017 No. 0527), and registered in the Chinese Clinical Trial Registry (https://www.chictr.org.cn/showproj.html?proj=32850).
Participant selection
All included patients met the following inclusion criteria. Patients were included if aged >59 till <81 years and diagnosed with unipolar or bipolar depression with or without psychotic symptoms according to the International Classification of Diseases, Tenth Revision criteria. Appropriateness for ECT was determined by a psychiatrist based on clinical indications. Typical reasons for referral to our center included multiple failed medication trials and illness severity/urgency. Other inclusion criteria were pretreatment 17-item Hamilton Depression Rating Scale (HAMD-17) total score >16 [17], ability to cooperate during the study, and capacity to provide voluntary written informed consent.
Exclusion criteria included: 1) a lifetime diagnosis of anesthesia risk, epilepsy, pneumonia, anesthetics allergies or fever, 2) patients participated in other related studies in 30 days, 3) patients accepted ECT or repetitive transcranial magnetic stimulation treatment in 6 months, 4) history of psychoactive substance abuse, and 5) patients with risk of cardiovascular disease.
Study design and ECT administration
During ECT, the antidepressant treatment plan was developed by a competent physician according to the patients’ clinical indications. For subjects who used benzodiazepine drugs, we will discontinue benzodiazepine drugs 2 days before the first treatment. For subjects accompanying psychiatric symptoms, antipsychotic drugs (11 subjects from msECT group taking antipsychotic drugs, 12 subjects from ECT group taking antipsychotic drugs) will be administered upon admission and the treatment dose will be gradually titrated. A total of 60 hospitalized patients with late life depression were randomly assigned receive msECT or RECT at a ratio of 1:1 using a computer-generated random number sequence. The number sequence list was placed in sealed envelopes and blinded to participants, and the neuropsychological measurement raters; it was only accessible to the psychiatrist administering ECT. The number of msECT or RECT sessions administered to the participant was guided by the progress and clinical improvement of the participant.
The participants were advised to fast 6 hours before anesthesia induction, which was performed using 0.5 mg intravenous atropine, propofol (general anesthetic), and succinylcholine (muscle relaxant). Once anesthetized, the participants were oxygenated throughout the procedure with 100% O2. Blood pressure, heart rate, and pulse oximetry were continuously monitored. To better ensure anesthesia effectiveness and reduce side anesthesia risk, the depth of anesthesia was monitored using the bispectral index (around 70). When the myoelectricity monitoring shows 4 consecutive stimuli as 0, perform modified electroconvulsive therapy operation. The American Somatics Thymatron System IV equipment (Somatics, LLC., Venice, FL, USA) was used for ECT, and the electrodes were placed according to the treatment standard of bitemporal [18] electric stimulation. At the first ECT a stimulation dose of 3.5 to 4 times age (in mC) (%energy or expected joules=70%–80% of age) is selected [19]. During the subsequent treatments (i.e., phase 2), the msECT group received lower stimulation frequency and stimulation time (5% energy, regardless of whether a seizure was induced) to reduce electrical dosage. During the subsequent treatments the RECT group received the same stimulation strategy regardless of reduction of the seizure duration during the course. The pulse width of the two groups was set at 1 ms during the entire treatment process.
Mood outcomes and side effects
A routine medical assessment was performed by a well-trained research team member for evaluating efficacy and side effects. The HAMD-17 was used to rate depressive symptoms at baseline, after the third session, after the last session, and 2 months after the last session. The impact of different stimulus doses on neurocognitive performance was measured using the Mini-Mental State Examination (MMSE) test prior to ECT, after the third session, and after the last session. Side effects were assessed as adverse events prior to ECT, after the third session, and after the last session. Participants with poor curative effect and those who did not complete 12 ECT sessions, regardless of the reason for discontinuation, were also followed-up till the end of the last session for final test measurements.
Primary and secondary outcomes
The primary outcome of efficacy was evaluated using the remission rate of depression, and the secondary outcomes included the number of ECT sessions required to achieve remission, improve neurocognitive performance, and reduce side effects. Nonresponse was defined as ≤25% reduction [20] in the HAMD-17, with the score being >7 points, whereas remission was defined as score ≤7 points. Participants who completed all treatment sessions but could not cooperate during evaluation were deemed lost to follow-up, and any failure to complete the two-stage treatment was defined as drop-out.
Statistical analyses
The full analysis set was used to analyze the study data. The intention to treat principle was adopted to analyze the main results. The remission rates of the two groups were compared using Pearson’s chi-squared test, and Dunnett’s (two-sided) t-test was used as the post-hoc test. The two-sample t-test was used to compare the number of ECT sessions required for remission. Moreover, analysis of variance was used to compare HAMD-17 and MMSE score differences at different time points within the group. The chi-squared test was used to compare frequencies of adverse events. The generalized estimation equation (GEE) method was used to evaluate the clinical efficacy and side effects between the groups over time. The working correlation matrix was set as the first stage of autoregression [21,22]. The treatment factors and visit time were considered as factors and age, equivalent dose of fluoxetine, and course of disease were used as covariates in the GEE of HAMD-17. In addition, the treatment factors and visit time were used as factors and age and number of ECT sessions were included as covariates in the GEE of MMSE. To identify whether msECT can induce seizures, the two sample t-test was used to compare the means of seizure duration and postictal suppression index (PSI) of the two groups. All statistical analyses were performed in SPSS 20.0 (IBM Corp., Armonk, NY, USA) for Windows. Effect size was calculated using practical meta-analysis effect size calculator (https://www.campbellcollaboration.org). For correlation analysis, Pearson’s correlation coefficient was run to measure a linear relationship between depression severity after the ECT course and age, number of sessions, sex, treatment modality, baseline depression severity. Used Pearson’s correlation coefficient to measure a linear relationship between cognition after the ECT course and age, number of sessions, sex, treatment modality, baseline cognition. We use multiple linear regression model to predict cognition after the ECT course.
RESULT
Patient characteristics
Out of 60 participants, one participant from the RECT group dropped out of the study at the third session owing to adverse events (behaving strangely); the results of the last visit of this participant was considered the treatment outcome of ECT according to the intention-to-treat principle. Other subjects completed their ECT course. Table 1 shows the basic demographic and treatment data.
Efficacy evaluation
Among those treated with msECT, 63.3% (19/30) met the remission criteria, 6.7% (2/30) did not respond (met the nonresponse criteria), and the mean decrease in the HAMD-17 score was 21.767 points, with a mean final score of 7.10 (standard deviation [SD]= 6.7). The same remission and nonresponse rates were observed in the RECT group, and among those who received RECT, the mean decrease in the HAMD-17 was 21.867 points, with a mean final score of 7.40 (SD=6.6). Among those in remission, the mean number of ECT sessions required for remission did not differ significantly between msECT (n=30) and RECT (n=30) (8.79 vs. 7.74, standardized mean difference [d]=0.4765, 95% confidence interval=0.1683 to 1.1214, p=0.15). Moreover, there was no interaction between the groups with respect to time (group×visit: Wald χ2=7.275, p=0.064), indicating that there was no significant difference between the two groups with respect to the extent of change in treatment efficacy over time (Figure 1A).
Overall cognitive function
The two different methods of ECT had varying effects on cognitive function during phase 2 (group×visit: Wald χ2=8.463, p=0.015). The estimated marginal means method was used to compare the trend of MMSE (Figure 1). In the first phase, the trend of MMSE in the two groups was similar. In the second phase, however, msECT showed an upward trend and RECT showed a downward trend (presented in Table 2 and Figure 1B). The mean recall subscores at the first two visit did not differ significantly for both group (visit one: ECT 2.43± 0.68, msECT 2.20±0.96, t=1.09, p=0.28; visit two: ECT 2.17±0.65, msECT 1.93±1.01, t=1.06, p=0.29). The mean recall subscores of ECT group at three was significantly lower than msECT group (ECT 1.77±0.77, msECT 2.23±0.94, t=-2.11, p=0.04).
Correlation Analysis and multiple linear regression
The relationship between depression severity after the ECT course and age (r=0.334, p=0.083), number of sessions (r=0.392, p=0.265), sex (r=0.088, p=0.502), baseline depression severity (r=0.033, p=0.815), treatment modality (r=-0.025, p=0.847) is not significant presented in Figure 2A-C. The relationship between cognition after the ECT course and age (r=-0.073, p=0.652), sex (r=-0.121, p=0.355) is not significant, but cognition after the ECT course show correlation for treatment modality (r=0.288, p=0.026), baseline cognition (r=0.457, p=0.038) (presented in Figure 2D-F). This multiple linear regression model (Table 3) expressed as in Eq. (1), where Y are dependent and X1 are baseline MMSE, X2 are treatment modality, X3 are number of ECT sessions.
Y=5.192+0.278X1+0.27X2+0.244X3. Eq. (1)
Comparison of epileptic seizures
The RECT group received a total of 90 sessions in the first phase and 232 sessions in the second phase, whereas the msECT group received 90 sessions in the first phase and 262 sessions in the second stage. The average duration of electroencephalography-confirmed seizure activity and PSI were shorter and lower, respectively, in the msECT group during the second phase (Table 4 and Figure 1). During the second phase of ECT, the average number of seizure duration ≥15 seconds [22] did not differ significantly between the RECT and msECT groups (3.93 [SD=2.080] vs. 3.00 [SD=1.702], d=0.4948, 95% confidence interval 0.0234 to 0.0129, p=0.630), whereas msECT resulted in a significantly lower rate of PSI ≥75% [23] (86.4% vs. 54.0%, p<0.001).
The seizure duration in both groups showed a downward trend. The duration of seizure in the RECT group showed a transient recovery at the 5th session, and the seizure duration in the msECT group showed a transient recovery at the 6th session. In the first two sessions (4th and 5th) after the reduction of electrical energy, the duration of seizure in the msECT group did not appear ideal compared with that in the RECT group, which may mislead the treating psychiatrist to terminate the reduced electrical energy strategy. However, in the 6th session, the average seizure duration of the msECT group was longer than that in the RECT group, suggesting that caution and patience are required after reducing energy in the second phase (Figure 1C).
During the second phase of ECT, the average number of seizure duration ≥15 seconds 22 did not differ significantly between the RECT and msECT groups (2.90 [SD=1.918] vs. 3.00 [SD=1.762], d=0.0547, 95% confidence interval 0.5652 to 0.4558, p=0.834), and there was no significant difference in PSI ≥75% between the two groups (63.5% vs. 53.0%, d=0.3781, 95% confidence interval 13.69% to 89.31%, p=0.153) (Figure 1D).
Side effects and safety
In the first phase, 32 adverse events affected 19 patients in the RECT group and 26 adverse events affected 21 patients in the msECT group. In the second phase, 145 adverse events affected 25 patients in the RECT group and 95 adverse events affected 16 patients in the msECT group. These events, which may be related to ECT, are shown in Table 5. In the first phase of treatment, one subject in the msECT group developed pathological Q-wave. No patient from either group died during treatment. Significantly fewer subjects in the msECT group suffered from adverse events in phase 2 (χ2=13.467, p=0.004).
DISCUSSION
The present study revealed that the electrical stimulation strategy of msECT did not result in reduced clinical efficacy. Moreover, compared with RECT, msECT minimized side effects in the second phase of treatment. In a multicenter randomized controlled experimental study, Kellner et al. [24] administered bitemporal ECT with 1.5 epilepsy threshold to patients (n=230) with depression aged 20–87 years and reported a total remission rate of 64% (95% confidence interval 53%–75%), which is not different from the remission rate observed in the present study. Regenold et al. [12] administered 1/8th dose of the standard regular ECT using the half-age strategy to 11 patients and achieved 73% final response and 55% remission rate. Regenold et al. [12] used the half-agesstrategy and bifrontal electrode placement for ECT, with a mean charge of 13.3 mc (SD=5.3) and fixed pulse width of 0.5 ms. By contrast, in the present study, the mean charge in the second phase was 30.014 mc (SD=39.00) and the fixed pulse width was 1 ms. Compared with the study results of ECT without convulsions, the present study revealed a certain remission advantage.
All subjects in the present study received drug treatment. Henkel et al. [25] reviewed the drug treatment of 1,014 patients with depression from multiple centers and suggested that early improvement was highly valuable at both visits on day 14 and 28. The meta-analysis conducted by Ottosson and Odeberg [26] suggested that as the antidepressant effect of ECT is sufficiently strong, the addition of an antidepressant agent is of minor practical value for ECT potentiation. These evidences suggest that the requirement for a combination of drug treatment with ECT for a subject may reduce treatment time required for remission. Moreover, the response time of drug treatment is slow. During the acute course of ECT combined with drugs, the clinical efficacy of ECT is more significant.
The present study results revealed that compared with RECT the average duration of seizure activity in the msECT group was shorter. However, whether there is a positive relationship between generalized seizure duration and response time remains unclear [26,27]. At present, it is believed that seizures lasting for 15–25 seconds is necessary for antidepressant treatment [22]. From the perspective of seizures, the msECT group experienced sufficient seizures (≥15 seconds) in both the entire treatment period and the second phase. Previous research has shown that the seizure threshold gradually increases with treatment progression [22]. The different seizure durations in the two study groups revealed that different energy levels had different effects on the seizure threshold. The decrease in the energy level appears to have different effects on the excitability of neurons, thereby delaying the transient recovery of seizure duration in the msECT group. This peculiar phenomenon deserves further exploration and may help clinicians understand the adaptive process of an individual to electrical stimulation.
msECT is less prone to produce effective seizures, but this drawback can be resolved by increasing the number of ECT sessions. msECT required an average of 8.79 sessions (SD=2.323) to result in a similar antidepression efficacy to RECT (which required 7.74 sessions), although this difference was not significant. It is known that ECT with higher energy requires fewer sessions for remission [16]. The first three standard ECT sessions may result in the requirement of fewer treatment sessions. D’Cunha et al. [28] reviewed the ECT of 221 elderly patients with depression in Australia, who received bitemporal ECT for an average of 9.1 (SD=4.5) sessions. They showed that the practice of reducing electrical energy in the later phase did not significantly increase the number of overall sessions. In the present study, there was no significant difference between the two groups with respect to the two phases of PSI. Postictal inhibition in msECT may be associated with an antidepressant effect by affecting the depression-related circuit in the thalamus [8].
In the second phase of treatment, msECT showed an upward trend in the overall cognitive level, indicating that the side effect of msECT on cognition is lower than RECT. Studies have reported decline in attention, memory, and executive function among elderly patients with depression [29,30]. Given that decreased depression can lead to improved performance in cognitive function [30], the complex effects of improved depression on cognition during the ECT may go undetected. The overall cognitive impairment and anterograde amnesia caused by ECT usually last for a short time [31]; retrograde amnesia usually lasts longer [32]. Therefore, the actual impact of reduced electrical energy on cognition could be offset by improvements in depression, which requires confirmation using the premorbid level of cognition as the control. Devanand et al. [33] reviewed clinical studies and animal research of imaging, anatomy, neurophysiology on ECT. They found no evidence of structural changes or neuronal loss associated with ECT [33,34]. Rami-Gonzalez et al. [34] placed a special emphasis on electrical and chemical changes involved in ECT-induced memory dysfunction. They argued that changes in electrochemical levels, abnormal functioning of the long-term potentiation process, and decreased cholinergic transmission contribute to transient memory deficits and incapacity in obtaining new information-after ECT [34]. An excessive release of excitatory amino acids and activation of their receptors (leading to oxidative stress) [35] as well as reduction in cholinergic transmission in the central nervous system [36], all of which are associated with ECT (particularly seizures), lead to adverse effects on cognitive functions. ECT causes a massive release of gamma-aminobutyric acid neurotransmitters and inhibits the activity of the cerebral cholinergic system [34]. Compared with RECT, the shorter duration of seizures in the msECT group may reduce side effects on cognition by reducing the oxidative stress and alleviating the decrease in transmitter levels. Besides, the applied stimulus intensity determines the magnitude of the electric fields [8]. Reduced stimulation intensity results in a smaller range of influence of electrical stimulation on brain regions which may be related to fewer adverse reactions from msECT.
Post-ECT delirium [37,38] can occur in up to 12% patients receiving ECT. Despite fewer post-ECT delirium events were reported among patients in the msECT group, it was difficult to conclude that msECT can reduce the incidence of delirium because the number of delirium events in the RECT group in the first stage of treatment was more than that in the msECT group. Jo et al. [39] reviewed 268 electronic medical records of patients who received bitemporal ECT and found no association between post-ECT delirium and ECT dosage. Thus, psychiatrists considering the use of msECT should balance delirium risk and therapeutic gain. Electrical and chemical changes during ECT participate not only in the occurrence of curative effect but also in the occurrence of side effects. Thus, finding a balance between the two is imperative in improving the efficacy of ECT.
At present, there are many ways to improve the efficacy of ECT, however, no form of this therapy can achieve 100% antidepressant effect. Sánchez et al. [22] emphasized that there are significant individual differences in ECT. Therefore, identifying the appropriate improvement method for a certain individual is extremely important. During the subsequent two msECT sessions, unsatisfactory epileptic seizures often dissuade psychiatrists from administering low energy stimuli when adverse reactions occur. However, the present study results support the use of msECT in clinical practice and suggest the possibility of lower cognition impairment. However, it is necessary to conduct additional randomized double-blind studies with equal sample sizes. Moreover, the mechanism of the fluctuating trend of seizure durations after the reduction of electrical energy need to be evaluated via imaging and other methods.
Limitation
Owing to the small sample size of this study, the statistical power may be lower. This may lead to a false negative result in sessions msECT therapy needed to cure depressive symptoms compared to conventional ECT. This monocenter, blinded one-sided design of this study may affect its external validity. Further, MMSE is not sufficiently sensitive for retrograde amnesia. Repeated measurement may lead to a bias in cognitive impairment that is not easily detected. Since this article uses a remission rate rather than a response rate to evaluate the antidepressant efficacy of electric stimulation therapy, the remission rate reported in this article may be lower than the response rate reported in other studies. Finally, similar to other studies, as the level of pre-disease cognition was unknown, the real effect of the second phase of msECT on cognition could not be accurately determined. This study indicates the potential of reducing the risk of adverse reactions by adjusting electricity consumption. In the future, it is necessary to expand the sample size and conduct further research using a double-blind randomized controlled trial.
