Rethinking the impact and management of electroconvulsive therapy session number in depression

Trajectory of Depressive Improvement

Depression is a common and disabling psychiatric disease accompanied by high suicide attempts (14, 15). Although ECT has significant advantages over antidepressants in efficacy and course, clinicians and patients seek clarity regarding the pace at which clinically meaningful benefits manifest, which defined as either achieving remission (i.e., an asymptomatic state) or response (i.e., a > 50% reduction in baseline symptoms severity), or reaching a plateau (i.e., no change in depression score). For example, what is the extent of depression improvement after each ECT session? when does the response or remission onset? What is the discrepancy in the speed of remission across different dimensions of depression? Which situations have a more rapid response following ECT?

Investigators have documented a nonlinear antidepressant response pattern over ECT. the Consortium for Research in ECT (CORE) conducted significant research in this field, gathering Hamilton Rating Scale for Depression (HRSD) scores after each ECT in 576 patients (16, 17). Their findings revealed a notable decrease in the mean HRSD score by 25.8% after the first session, 39% after the second session, and 49.3% after the third sessions. It was observed that the median time to first response was typically three ECT sessions, with remission achieved after approximately four additional ECT sessions, demonstrating an early improvement trajectory. Other studies have similarly reported rapid effects of ECT in comparable populations. In the Prolonging Remission in Depressed Elderly (PRIDE), involving 185 geriatric depressed patients received right unilateral ECT, the mean decrease of HRSD scores in the first three ECT sessions were 24.5%, 35%, and 42.7%, respectively (18). Additionally, Rodger et al. reported the change in HRSD score between first and third session was six times greater than the remaining sessions (19). Overall, the trajectory of depressive improvement during ECT appears to be swift in the early stages, leveling off in the later stage, reflecting the relatively rapid response to ECT.

While average trajectories offer valuable insights, substantial individual differences remain in the speed of response to ECT. Clinical case reports have described varied response patterns, including rapid response after one session and delayed improvement following 10 sessions (20, 21). A large prospective cohort study reported that 12.6% of patients responded after the first session, whereas 5.9% showed no response throughout the treatment course (16). Another study found that 40% patients recovered with two to four ECT sessions, 40% with five to eight sessions, and only 20% required nine to 12 sessions (22). To better characterize the variability, researchers have applied data-driven methods to identify distinct response subgroups. Latent class analysis of 156 consecutive patients identified five distinct trajectories of depressive symptoms, including rapid improvement (25%), moderate improvement (30.12%), slow improvement (19.23%), slow improvement with delayed onset (11.54%), and no improvement (12.82%) (23). Similarly, growth mixture modeling in 239 patients identified three patient groups consisting of rapid response group (16.74%), slow response group (76.15%), and nonremit group (7.11%) (24). Taken together, these studies underscore the heterogeneity in ECT response trajectories. Recognizing and accounting for these variations may help guide individualized treatment duration, optimize outcomes, and reduce unnecessary exposure to prolonged ECT courses.

As widely acknowledged, depression is a heterogeneous disease characterized by multiple distinct symptom clusters, including mood, anxiety, somatic, insomnia symptoms, and suicidal ideation, among others (25). When treating depression, it is essential to recognize that not all symptoms improve at the same rate or degree (26). Relying solely on total scores of depressive symptom severity to define responses may lack detection of resolved and residual symptoms. Some studies have suggested that suicidal ideation respond quickly to ECT (27), hence proposing suicide risk as an indication for ECT. In contrast, a study involving 89 older persons with depression found that while all dimensions showed rapid and significant improvement, the mood dimension demonstrated the highest rate of improvement compared to suicidal dimensions (28). These results are consistent due to the differing proportions of each dimension in the HRSD scale; for example, the suicidal dimension comprised only one item. Considering the covariation of symptoms over time, a dynamic time warping analysis of 68 participants showed that improvements in somatic symptoms and suicidal ideation preceded those in mood symptoms (29). In conclusion, the temporal trajectories of symptom clusters vary, and it remains debatable which sets of symptoms are most effectively and rapidly targeted.

In addition to subgroup analyses of response trajectories, several studies have investigated predictors of early and delayed responses to ECT. Notably, patients with bipolar depression, psychotic features, and higher depression severity at baseline showed a more rapid response after ECT (24, 30). There is evidence that among initial responders, patients with unipolar depression require an average of six treatments to meet the response criteria. In contrast, patients with bipolar depression meet the response criteria after four treatments (30). In the CORE study, patients with psychosis had an average percentage change in HRSD of 64% compared to 56% for nonpsychotic groups after the fifth ECT (31). Additionally, a regression model demonstrated that baseline HRSD scores were significantly associated with a rapid response. Other clinical characteristics also affect response speed. Treatment-resistant depression (TRD) is often associated with slower improvement, while comorbid personality disorders are linked to a higher likelihood of nonresponse (23). In contrast, first-episode depression does not appear to significantly influence response speed (23). The role of age in response speed is inconsistent and complex. While some studies indicate that elderly patients experience faster remission than younger patients (32), another study reported that elderly patients with depression require more ECT treatments than adults (33). This discrepancy arises from various factors interfering with the analysis of the independent role of age. Regarding the ECT technique, response speed is associated with electrode placement (34). Kellner et al. observed a decrease of 44% in HRSD scores after the first session of right unilateral ECT, 48% for bifrontal ECT, and 51% for bitemporal ECT (35). Despite growing insights into potential predictors, systematic evidence on the temporal dynamics of ECT response remains scarce, as most meta-analyses emphasize overall outcomes (36, 37). Future studies should investigate how clinical and technical factors shape response trajectories, to support more personalized and effective ECT protocols.

Trajectory of Cognitive Impairment

Cognitive impairment is a frequent adverse effect of ECT among patients, which can be divided into memory-related and nonmemory cognitive impairment (38). Memory-related issues include disorientation, anterograde amnesia and retrograde amnesia. Nonmemory cognitive impairment involves decreased attention, processing speed, and executive function (39). National guidelines recommend the cognitive impact of ECT should be monitored on an ongoing basis. However, in clinical practice, key questions remain unanswered: when does the cognitive impairment occur? Does the trajectory of cognitive impairment worsen or improve over time? What is the relationship between depressive symptoms and cognitive impairment?

Cognitive deficits emerge early in ECT (40). A prospective follow-up study found that 62% of patients with depression reported subjective memory deficits after the first ECT session, a figure that increased with subsequent ECT sessions (41). Although these deficits typically resolve after all modified ECT sessions are completed, 34% of the deficits may last for 6 months or longer (42). Notably, disorientation commonly surfaced immediately after the second ECT session and was more pronounced after the fifth ECT session (43). However, these results are somewhat subjective due to the nonspecific learning effects associated with cognitive measurement tools, hindering accurate assessment post-ECT. To address these limitations, some researchers have adopted strategies such as employing multiple parallel sets of tests or reducing the frequency of measurements. Viswanath et al. applied a short cognitive-related battery in 30 inpatients and they found that objective cognitive deficits such as verbal memory, autobiographic memory, and psychomotor speed progressively deteriorated from the first to the third to the sixth ECT session (44). However, this study lacked a baseline assessment. Another study compared a new electroconvulsive therapy cognitive assessment (ECCA) tool with the classic Montreal Cognitive Assessment (MoCA). It indicated that ECCA scores were significantly decreased across the three testing points, whereas MoCA scores did not vary significantly (45). An intensive longitudinal follow-up of associative memory at five timepoints found that memory impairment occurred after the first ECT and worsened during subsequent ECT treatments (46). It is important to note that cognitive deficits are influenced by factors such as age, education level, and medications, including anesthetics and antidepressants. For instance, older individuals with lower education levels tend to experience more severe cognitive impairment (46), while substances like propofol, low-dose ketamine, and lithium may have potential cognitive-protective effects (47–49). In summary, cognitive impairment occurs early during ECT and may accumulate throughout the treatment course, implying the importance of understanding the optimal number of ECT sessions to mitigate unnecessary cognitive damage.

Emotion and cognition constitute the two main elements of neuropsychology (50). Despite ECT induces rapid improvement in depressive symptoms alongside cumulative cognitive impairment, the relationship between depression and cognitive function remains unclear. Clinical perspectives suggest that cognitive impairment might aid depressive remission by enabling patients to forget distressing memories (51). Bai et al. found that negative memory impairment was more severe than positive memory impairment and correlated with symptom relief (10). Conversely, depressive remission can enhance certain cognitive functions, such as increased subjective initiative (39, 52). From a neurophysiological perspective, the seizures induced by ECT, especially those aimed at enhancing efficacy, are often linked to cognitive side effects. For instance, compared to ultra-brief pulse width, brief pulse width has proven to be more efficacious regarding symptom reduction but resulting in more pronounced cognitive side effects (53). Right unilateral ECT typically yields milder and less persistent cognitive effects but slower response rates compared to bilateral ECT due to reduced stimulation of the left temporal lobe (54). Additional research suggests that hippocampal changes caused by ECT are involved in both depressive improvement and cognitive impairment (55). In conclusion, while emotion and cognition are closely connected, the exact nature of this relationship, the role of epileptic seizures, and the whole-brain alterations need further investigation. Before clarifying these questions, regular monitoring of both depressive symptoms and cognitive function during ECT is essential to help clinicians decide the optimal time to end treatment.

Trajectories of Hematological and Neuroimaging Markers

Understanding the mechanisms of treatment responses is paramount for improving depression outcomes. However, due to its spatially unfocused nature, the neural mechanisms underlying the clinical response to ECT remain uncertain (68). Various hypotheses have been proposed to explain the effect of ECT, including neurotransmitter, neuroendocrine, neuroinflammation, and neuroplastic changes (12). For instance, monoamine neurotransmitter systems, such as norepinephrine, and inhibitory neurotransmitter systems, such as gamma-aminobutyric acid (GABA), have been discussed as potential mediators of therapeutic response in ECT (11). Additionally, ECT triggers the release of neurotrophic factors, adrenocorticotrophic hormones, and inflammatory mediators, including interleukin-6, and cortisol (69). Moreover, ECT brings about widespread changes in the structure and function of the brain, attributed to neuroplastic effects (70). Although these effects are easy to detect, challenges persist in research on ECT-induced neurological effects. For example, questions arise regarding the temporal relationship between these neurological changes following ECT and the treatment effects. Moreover, there is a need to understand the internal relationship among these factors and determine which of these changes may be related to the antidepressant and amnesic effects or incidental phenomena.

Most existing studies have investigated neural alterations before and after ECT, which is the total effect of multiple ECT sessions, resulting in key changes being masked (71, 72). However, few studies have focused on the time course and processes of neural effects during ECT treatment. Regarding hematological markers, a longitudinal study spanning nine visits during the ECT period indicated that serum brain-derived neurotrophic factor (BDNF) levels increased after each ECT session yet showed no significant association with treatment response (73). Similarly, an exploratory study evaluated the acute endocrine effects and found that levels of cortisol and norepinephrine were significantly elevated after the first ECT session and fall back to baseline after the course of ECT (74). Another study by Göteson et al. investigated alterations in the serum proteome of 309 patients before and after the first ECT session and before the sixth ECT session, revealing findings related to signal transduction; however, none of the studied protein biomarkers were associated with the clinical response to ECT (75). In addition, no significant changes were found in white blood cell, proinflammatory cytokine/neurotrophin ratios, and plasma vascular endothelial growth factor levels over the course of ECT (76). In essence, on a finer time scale, the hematological markers associated with nutritional factors, inflammation, and endocrine function exhibit transient surges, potentially attributed to the stress induced by ECT.

In particular, because MRI is relatively safe without ionizing radiation, it has enabled repeated scanning of patients at various intervals to track the longitudinal trajectories of structural and functional cerebral changes during ECT. The majority of imaging studies are acquired before, mid and after treatment and focused on the hippocampus and the amygdala. Shantanu et al. scanned 43 patients with major depression at three timepoints: before ECT, after the second session, and within 1 week of completing the ECT series, and found progressive increases in hippocampal and amygdala volumes, which were associated with symptom improvement (7). Smaller baseline hippocampal volume predicted greater clinical response. Similarly, another longitudinal study of 14 patients, with MRI assessments conducted before ECT, after the fifth or sixth session, and at treatment completion, demonstrated a trend toward increased hippocampal volume across sessions. A large association analysis between MRI data and the number of ECT sessions from the Global ECT-MRI Research Collaboration (GEMRIC) reported a 0.28% linear increase in hippocampal volume after each ECT session (77–79). Marta et al. further examined hippocampal metabolites and amygdala functional connectivity across an acute course of bitemporal ECT including pretreatment, after the first and ninth sessions, and 15 days posttreatment, identifying sequential changes in neuroinflammatory markers and limbic network activity (80). These findings support the role of ECT-induced neuroplasticity in the hippocampus and amygdala in mediating clinical improvement in depression. Beyond the hippocampus–amygdala complex, longitudinal changes of gray matter volume or cortical thickness in the thalamus, putamen, and anterior cingulate cortex, as well as fractional amplitude of low-frequency fluctuations in the subgenual cingulate cortex and activation intensities within auditory networks have also been reported during ECT course. A summary of the longitudinal neuroplastic effects of ECT is provided in Table 1. Taken together, ECT appears to affect a broad range of brain regions, aligning with the distribution of electric field strength; however, its core neurobiological mechanisms remain unresolved (81). Moreover, findings from longitudinal studies suggest that ECT may induce brain changes at an early stage, consistent with the rapid clinical response. Nevertheless, the dynamic impact of ECT sessions on brain structure and function remains unclear, as most studies include only a limited number of observation timepoints, typically three or four. Notably, although longitudinal designs rely on self-comparisons, stimulation parameters, medications, and individual traits may still confound imaging findings, yet their effects remain insufficiently studied due to limited data. Therefore, further longitudinal studies using multidimensional imaging markers across more timepoints or even throughout ECT while controlling for potential confounding factors, are needed to generate a comprehensive profile of neuroplastic changes over time and their relationship to therapeutic outcomes (80–86).

The mechanism behind the early and rapid response to ECT remains elusive, with the microscopic changes in the brain that accompany these responses are still under speculation. Studies in rodents suggest that even a single ECT session can have profound effects, including the activation of the immune system and neurotransmitter increase. Hippocampal neurogenesis, which is thought to be involved in the therapeutic effects of ECT, was significantly increased in single electroconvulsive seizures in a rat model, consistent with the increased hippocampal volume shown by neuroimaging (87). However, some researchers argue that the induction of neurogenesis or an increase in gray matter volume takes time and may not occur acutely (20). Integrating insights from micro-level biology, meso-level imaging, and macro-cognitive levels at multiple timepoints could provide a clearer understanding of how the number of ECT sessions affect outcomes and further clarify the rapid onset mechanism of ECT.

ECT Response–guided Sequential Strategy

Although multiple treatment options are available for depression, no single treatment approach can fully and safely address the complexity of the disorder (13). In recent years, sequential treatment strategies, which combine different interventions in a staged and adaptive manner, have gained increasing attention (88). This approach aims to maximize the benefits of each treatment at a specific stage while minimizing the side effects and risk of treatment resistance that can develop from long-term use of a single treatment modality (89, 90). Therefore, a sequential treatment strategy is a meaningful shift in clinical thinking, allowing the selection of appropriate alternative treatment options based on the patient's response to the first course of treatment (91). In an ideal scenario, the primary objective of an initial ECT course is to treat the current episode while minimizing cognitive impairment. The number of sessions plays an important role in the clinical effects of ECT. Through our literature review, we have identified a discernible pattern in both depression alleviation and memory impairment throughout the course of ECT. Initially, depression tends to improve rapidly with successive ECT sessions, followed by a slower rate of improvement, while memory impairment tends to accumulate gradually. This supports the view that it may not be necessary to perform ECT so many times. Therefore, managing ECT sessions and sequencing to other treatments may be an effective way to improve outcomes (88, 92). The notable advantage of ECT lies in its swift antidepressant effect during the early stages, which is unmatched by other treatment modalities. However, as treatment progresses, the cognitive and economic burdens tend to escalate (93). In light of these considerations, we propose a new treatment strategy in which ECT should be terminated early once the optimal benefit-to-risk ratio is achieved, and then the patient should be transitioned to other safer treatments rather than persisting until complete remission. Notably, the response trajectories of ECT vary among patients with depression. Therefore, we outline an ECT response–guided sequential strategy to provide tailored sequential treatments for different response trajectories (Figure 4).

View Figure

• Download as Powerpoint
• Download Figure

Based on the previously described speed of improvement and treatment outcomes, patients undergoing ECT can generally be categorized into three clusters: rapid responders with remission, slow responders without remission, and nonresponders. Rapid responders with remission are those who improve quickly in the early stages of ECT. Although continued ECT can lead to remission, it may increase the risk of memory impairment. For these patients, sequential use of portable and safe noninvasive brain stimulation (NIBS) or lithium after an ECT-induced response helps consolidate the treatment effects while reducing cognitive damage (94). NIBS is an emerging therapy that modulates neuronal activity through physical stimulation, such as electricity, with the advantages of safety, portability, precise regulation, and cognitive enhancement, which is recommended by the FDA for the treatment of depression (95, 96). Sequential application of NIBS after ECT may not only further alleviate depressive symptoms through targeted modulation of specific brain regions, but also help prevent cognitive impairment caused by excessive ECT, protect cognition-related brain areas, promote cognitive recovery, and provide a home-based option that is more acceptable to patients (97–99). An adequately powered exploratory efficacy study is currently underway to provide definitive evidence of NIBS following ECT (100). Slow responders without remission are those who show a slow response in the early stages of ECT and experience no further clinical improvement in the later stages. This may be due to poor neural plasticity or insufficient precision of ECT (101). Augmentation strategies that lengthen seizure duration, such as applying bilateral ECT or increasing the electrical dosage, may improve efficacy, but at the expense of cognitive impairment (58). Alternatively, switching strategies may be considered, replacing ECT with fast-acting treatments such as ketamine, an N-methyl-D-aspartate antagonist, which has demonstrated rapid antidepressant effects and has been found to be noninferior to ECT (102). In addition, supplementation strategy using personalized neuromodulation therapies shows considerable promise in addressing individualized residual symptoms that may persist after ECT, such as anhedonia or somatic complaints. These symptoms are often not fully resolved by standard ECT protocols and may require targeted interventions tailored to specific neural circuits or symptom clusters, thereby complementing the antidepressant effects of ECT and enhancing overall recovery (103). The third group, a small subset of patients, showed minimal clinical response to ECT and were labeled nonresponders. The reason for poor efficacy is often due to the co-occurrence of personality disorders, making these patients more suitable for targeted psychotherapy or thought training (104). In summary, this response-guided sequential treatment strategy consists of two phases: an initial phase leveraging the rapid antidepressant effects of ECT, followed by tailored follow-up interventions based on individual response trajectories, such as maintenance, augmentation, switching, or supplementation, to optimize clinical outcomes while minimizing cognitive burden and overtreatment.

Although the theoretical foundation and indirect evidence supporting the sequential treatment strategy are compelling, further rigorous testing in future studies is essential. A key challenge lies in accurately characterizing individual ECT response trajectories and determining optimal sequential treatment approaches. Emerging evidence suggests that baseline predictors, including clinical characteristics (e.g., age, depression subtype, treatment resistance) and neuroimaging markers (e.g., hippocampal volume), may help forecast these trajectories and subsequent treatments, with further refinement of predictions achievable by integrating dynamic data, such as symptom reduction rates and biological changes during early treatment sessions (105). However, the reliable biomarkers and clinical features to guide transition decisions require investigation and validation in future studies. Another critical issue is determining the appropriate time to transition from ECT to subsequent interventions. Ideally, the cut-off point for ECT should be based on achieving optimal levels of neuroplasticity and disruption, providing a sufficient antidepressant response with minimal side effects (70). On average, responsive patients may require 3 to 4 ECT sessions to reach this critical point, while nonresponders should discontinue ECT, as it may no longer be suitable for them. It is important to note that this sequential strategy primarily addresses the acute treatment phase and does not replace the need for long-term consolidation, which is typically maintained with pharmacotherapy. Nevertheless, tailoring sequential treatment strategies based on ECT response trajectories holds significant potential for improving clinical outcomes, particularly by offering greater flexibility and safety in addressing challenging conditions such as TRD. We strongly encourage future clinical trials to further explore and validate this approach, ultimately enhancing understanding and improving treatment outcomes.