The centenarian population is growing significantly and is projected to increase sharply in the coming decades. As of 2024, there were approximately 722,000 centenarians worldwide, with Japan, the United States, and China having the highest numbers [1]. This figure is expected to rise to 3.4 million by 2050 and exceed 25 million by 2100 [2,3]. In Europe, notable examples include France, Italy, and Greece, each with >20 centenarians per 100,000 inhabitants registered already in 2011. Portugal also stands out with a density of 14.4 centenarians per 100,000 inhabitants that same year, ranking it among the top 15 in Europe [4]. In Switzerland, the centenarian population more than doubled from 787 in 2000 to 1948 in 2022, reaching a noteworthy prevalence of 22.1 per 100,000 inhabitants [5].

Amid extreme longevity, individuals’ health experiences vary significantly, from robust health to substantial declines [6-8]. Depression, a prevalent mood disorder among older adults [9], emerges as a major public health concern [10]. The International Classification of Diseases, 11th Revision characterizes a depressive episode as “a period of depressed mood or diminished interest in activities occurring most of the day, nearly every day during a period lasting at least two weeks accompanied by other symptoms such as difficulty concentrating, feelings of worthlessness or excessive or inappropriate guilt, hopelessness, recurrent thoughts of death or suicide, changes in appetite or sleep, psychomotor agitation or retardation, and reduced energy or fatigue” [6]. However, this definition broadly encompasses the general population and may not fully address the possible difference and heterogeneity of depressive symptomatology in older adults, potentially leading to diagnostic imprecision.

Accurate diagnosis is crucial as depression should not be misconstrued as an inherent aspect of aging even in very advanced age [11,12]. A systematic literature review reported substantial variations in depression prevalence among near-centenarians and centenarians across countries, such as 12.8% in Italy, 13.5% in Australia, 20% in the United States, and 29% in Mexico [13]. These differences may reflect cultural, health care, and socioeconomic influences as well as methodological variations in study conduct [13,14]. Interestingly, comparisons between centenarian cohorts and younger groups have shown inconsistent findings, with half of the studies reporting higher rates of depressive symptoms among centenarians [13]. This highlights the diversity of aging experiences and suggests that mental health may not necessarily worsen with age. Particularly noteworthy are the findings from Sweden, where near-centenarians and centenarians exhibited almost double the depression rates of older adults aged 85 years at 32.3% versus 16.8%, respectively. These statistics highlight the need not only for careful mental health evaluations among individuals of very old age but also for prompt discussion on tailored approaches to support mental well-being in this population [13,15].

Depression can lead to severe physical, cognitive, and psychological effects. Physically, it can cause malnutrition, falls, delirium, functional decline, and increased mortality [13,16-20]. Cognitively, depression has been connected to cognitive decline and dementia [21-25]. Psychologically, it has been associated with anxiety and suicidal ideation [26,27]. Furthermore, depression places a significant burden on caregivers and health care systems [16-18,28,29], increasing costs and resource demands. These consequences underscore the critical need for early detection and intervention. Therefore, proactive measures such as regular mental health screenings are essential to prevent the severe impact of depression on individuals’ quality of life and alleviate its wider societal effects [11]; however, these are not part of the current practice in most countries worldwide.

Addressing this gap in care is difficult as depression in older adults frequently remains underdiagnosed and undertreated, in part because its manifestations are often misinterpreted as natural aspects of aging [18,19,30,31]. Studies have estimated that approximately half of depression cases are not identified by frontline health care providers [16,18], and of those detected, approximately half do not receive adequate treatment [16]. The challenge in detecting late-life depression (LLD) stems from its sometimes subtle, hidden, or atypical signs and symptoms [30,32]. Physical manifestations such as weight loss, psychomotor retardation, and exhaustion, as well as emotional manifestations such as loss of interest in activities, heightened anxiety, or irritability, might overshadow the typical sadness often associated with depression [9,18]. This complexity requires greater clinical vigilance. Moreover, the possible overlap of symptoms between depression and dementia adds another layer of diagnostic difficulty as both conditions can present with symptoms such as memory impairment, psychomotor retardation, and reduced motivation [9]. Beyond dementia, several other medical conditions can mimic or exacerbate depressive symptoms. Chronic pain, thyroid dysfunction, Parkinson disease, stroke, and medication side effects—particularly in polymedicated individuals—can contribute to depression-like symptoms, including fatigue, apathy, functional dependence, memory impairment, and disturbances in mood and sleep [33-38]. These overlapping symptoms underscore the need for a comprehensive assessment to distinguish LLD from other conditions, aiming for an accurate differential diagnosis. The wide range of signs and symptoms associated with LLD can extend beyond the conventional diagnostic criteria for depressive disorders. Symptom clusters not fully meeting the diagnostic criteria for depression may still carry significant clinical importance due to their link to reduced quality of life and increased disability [9,18].

Given the prevalence of depression among individuals of very old age, better detection of early signs and symptoms is crucial. However, despite the availability of various screening tools, to the best of our knowledge, none have been specifically validated for near-centenarians and centenarians (aged ≥95 years). Therefore, our study aimed to compile a comprehensive range of potentially relevant depressive features that have been documented in scientific literature. These features were then submitted to an international panel of experts to achieve consensus on the most representative signs and symptoms of depression in individuals of very old age. The outcomes of this approach are 2-fold. First, by systematically examining and integrating diverse diagnostic criteria and screening tools, this study aimed to contribute to the ongoing debate about the complexities and challenges of effectively diagnosing depression in individuals of very old age. Second, achieving expert consensus is central to developing a new screening instrument specifically tailored to this unique population. This effort might be an important first step toward early detection and effective intervention, which might lead to better outcomes and enhanced general well-being for this age group, and seeks to contribute to promoting mental health and supporting optimal care for individuals of very old age.

We conducted a modified e-Delphi study to systematically combine expert opinions and achieve an informed group consensus on the symptomatology of depression in very old age. The Delphi technique is an established approach in which a panel of experts is asked to provide their opinions on a given issue over the course of several rounds [39]. In each round, the questions are informed by the findings of the previous one, allowing the study to evolve over time based on the collected data [39].

Given the prevalence of depression among older adults and the complexities of its diagnosis, the Delphi technique is particularly useful for gathering expert consensus on this issue. In our study, we used an e-Delphi approach to involve experts from different locations and facilitate efficient data collection through electronic communication methods. To adapt the traditional Delphi approach to the specific requirements of our study, we implemented several modifications (referred to as a modified e-Delphi study). To keep the focus on the items rather than on the degree of agreement, experts were informed at the start of each new round about which items had reached consensus without being given the precise consensus rates. In addition, for items on which consensus had already been reached, experts were unable to review their previous responses. This decision was made to avoid revisiting settled issues and manage the large number of items efficiently, ensuring the stability of agreed-upon elements. An additional aspect of our modified methodology was a final evaluation of open issues by the steering committee. This last step took place to discuss the items that had not achieved consensus over the e-Delphi rounds, ensuring a thorough examination of these elements.

The materials presented to the e-Delphi panel were based on a comprehensive analysis of the existing literature on depression in very old age (personal communication by Gomes da Rocha et al, 2023). This ensured that the expert panel’s input was based on a thorough understanding of the relevant literature, thereby enhancing the validity of the final consensus, which may be seen as a lever to encourage scientific debate and foster new developments toward an effective diagnosis of depression in very old age [40].

This study adhered to the Conducting and Reporting Delphi Studies guidelines [41,42].

The steering committee for this e-Delphi study based in Western Switzerland comprised 4 members: 2 physicians who are old age psychiatrists and senior researchers (AvG and PV), 1 clinical nurse specialist in geriatric care and a senior researcher skilled in the e-Delphi technique (HV), and 1 clinical nurse specialist in geriatric care who is also a junior researcher (CGdR). The committee was responsible for overseeing the study design, ensuring ethical compliance, and guiding the iterative process of questionnaire refinement and consensus building. Meetings were held monthly, with methodological decisions made based on discussion. The committee played a crucial role in maintaining the study’s rigor, managing data collection and analysis, and addressing any lack of consensus among panelists.

In the Delphi approach, nonprobability sampling methods are used through the purposive selection of an expert panel. However, clear selection criteria were applied to limit researcher bias based on a five-step procedure [43]:

The Delphi method offers no consensus on an ideal sample size as the focus is on qualitative depth and the use of expert judgment for a comprehensive exploration of specific topics rather than on statistical representativeness [44]. Usually, a panel comprises 7 to 15 experts to facilitate effective information processing and decision-making; however, some studies may involve up to 30 experts for broader perspectives [44]. For our study, we aimed to assemble a convenience sample of at least 20 experts from various countries.

On the basis of the predefined selection of potential panelists, we emailed invitation letters to a total of 84 experts. These letters outlined the study’s objectives, described the e-Delphi context, detailed the questionnaire (eg, language, type of questions, and estimated completion time), and clarified what was expected from the participants, ensuring that they understood the significance of their contributions to the study. Interested parties could access the questionnaire through a direct URL to the survey. Anticipating a 25% to 30% response rate for the first-round questionnaire, we expected approximately 20 to 25 experts to participate. We aimed to retain these experts across all subsequent rounds, planning for a maximum of 4 rounds.

The SurveyMonkey tool (SurveyMonkey Inc) was used as the survey software as it facilitates secure and efficient data collection. In accordance with literature recommendations [43], each e-Delphi round remained open for a minimum of 4 weeks, during which up to 3 automatic reminder emails were dispatched to nonrespondents. At the start of each new round, experts were provided access to the overall results from the previous round, including a list of items for which consensus was either achieved (≥70% agreement) or not achieved (<70% agreement) [43]. The items that did not reach consensus were integrated into the current round for further deliberation, along with any new items that emerged from panel suggestions through open-ended questions.

The questionnaire was divided into 3 sections.

Before launching round 1, the questionnaire was pretested with a small purposive sample of 5 professionals (nurses, psychologists, and a physician) to verify its clarity and comprehensibility, with only a minor adjustment made based on 1 expert’s suggestion to include specific examples for certain items. Despite the positive outcomes of the pretest, experts were invited to rate the questionnaire at the end of round 1 for comprehensiveness, readability, and clarity.

Overall, 93% (26/28) of the experts participating in round 1 rated the questionnaire as Comprehensive or Very comprehensive, and 86% (24/28) found it Easy or Very easy to read and understand. Due to this positive feedback, no further structural adjustments were implemented before proceeding to the subsequent rounds.

Data collected via SurveyMonkey were transferred to a Microsoft Excel spreadsheet (Microsoft Corp) and examined for any potential errors or missing data. They were then imported into the SPSS Statistics software (version 29.0; IBM Corp) for analysis [58]. Descriptive statistics were computed to characterize the panel from a sociodemographic and professional perspective using means and SDs as well as medians and interquartile ranges for quantitative variables and frequency distributions and percentages for qualitative variables. The level of agreement among experts required to achieve consensus on a given item was set at 70% for each round. This threshold is commonly used in Delphi studies [59-61] as it allows for the inclusion of minority viewpoints in the early rounds, ensuring that diverse perspectives are considered before achieving final consensus. This is particularly relevant in the context of depressive symptom screening given the challenges associated with detecting LLD, as previously mentioned. Hence, this consensus rate is considered acceptable for preserving methodological rigor, as supported by the literature [43], which indicates that such a threshold aligns with norms for ensuring reliability and validity in the context of an iterative, consensus-building approach. The consensus rate for each item was calculated using a 4-level Likert scale (1=Not relevant, 2=Somewhat relevant, 3=Relevant, and 4=Very relevant). In the first step, these responses were recoded into 2 levels: values of 1 and 2 were recoded as “0” (Overall not relevant), and values of 3 and 4 were recoded as “1” (Overall relevant). Next, we aggregated the recoded responses for each item to determine consensus, which was achieved when the percentage of experts identifying an item as Overall relevant or Overall not relevant met or exceeded the predefined threshold of 70%.

After the first e-Delphi round, items that did not reach the consensus level were carried forward to the subsequent round. At the beginning of each subsequent round, experts were given the opportunity to review the general outcomes from the previous round (ie, consensus achieved or consensus not achieved for each evaluated item). Items that achieved consensus were not included again in subsequent rounds. After conducting 3 rounds with the participating panel and observing only limited changes in the consensus rate after the third round, we decided to proceed with the last round, which involved a final evaluation of unresolved aspects by the steering committee. This last step aimed to gain a deeper understanding of the reasoning behind the outcomes and further refine the consensus process. Items that still did not achieve consensus after the 3 e-Delphi rounds were thoroughly examined during this phase.

The Ethics Commission on Human Research of the Canton of Vaud was consulted and confirmed that this project fell outside the scope of the Swiss Human Research Act (Req-2023-00844; Multimedia Appendix 2) as the study did not involve the collection of health-related personal data or biological material.

All participants received detailed information about the study, including its aim and relevance, what to expect, a description of the questionnaire, instructions for participation, and contact information for the main researcher, along with the invitation email. The questionnaires for all rounds (1, 2, and 3) included an introductory section emphasizing the voluntary nature of participation in the study, as well as specific information regarding confidentiality, data protection, risks, and benefits and a statement on ethical considerations. Participants were explicitly informed that, by answering the questionnaire, they were implicitly providing consent to take part in the study. They did not receive any compensation for their participation. Each participant received a survey link for accessing the questionnaire, and their IP addresses were not recorded to ensure confidentiality. The collected data were deidentified, with no personal identifiers retained. Responses stored on SurveyMonkey were deleted after transferring the data to the Microsoft Excel spreadsheet. All data are securely stored on a password-protected institutional server to comply with Swiss data protection regulations and will be definitively deleted after 5 years.

A total of 84 experts were invited to this modified e-Delphi study. Of these 84 experts, 28 (33%) participated in the first round, forming the baseline panel. This group (n=28) was consequently invited to the second and third rounds, for which the participation rates were 75% (21/28) and 71% (20/28), respectively.

The panel exhibited sociodemographic and professional diversity, including different nationalities, professions, and fields of practice. The average age ranged slightly between 44.1 (SD 10.6) and 46.7 (SD 11.4) years over the rounds. A majority of women participated in the first and second rounds (17/28, 61% and 13/21, 62%, respectively), whereas men were more prevalent in the third round (11/20, 55%). Swiss and Portuguese nationalities were the most represented (12/20, 60% to 18/28, 64%), with the nursing profession being predominant (9/20, 45% to 12/21, 57%), followed by the medical profession (5/21, 24% to 8/20, 40%). The average professional experience was 19.4 (SD 8-9) years, with nearly 80% (22/28, 79%) of the experts being engaged in purely clinical activities or a combination of clinical and academic activities, among others (eg, service organization). More than 80% (23/28, 82%) reported having regular or very regular contact (daily, weekly, or monthly) with individuals aged ≥85 years, and 75% (21/28) reported having regular or very regular contact with individuals aged ≥95 years (Table 1).

The completion dates and the mean time taken for participants to complete the questionnaires in each round are summarized in Table 2. Globally, the e-Delphi rounds took place between August 2023 and February 2024, with an overall weighted mean completion time of 16 minutes and 40 seconds (SD 11 min and 27 s).

A consensus threshold of ≥70% was established for both dimensions and criteria, as previously described.

In round 1, experts (n=28) were asked about the importance of considering the duration of depressive signs or symptoms in addition to their presence. It is important to note that, for the purpose of this assessment, they were instructed not to consider suicidal thoughts or ideation as these require immediate attention and, thus, were deliberately excluded from the duration criteria. All participants (28/28, 100%) agreed on its importance. When specifically asked about the minimum duration for considering these signs or symptoms as clinically significant, 25 experts provided their insights—21 (84%) suggested a range from a few days (1 to 3) to 4 weeks, with 2 weeks being the most cited (n=9, 36%).

In addition, 7% (2/28) of the experts suggested differentiating the duration of signs or symptoms between recurrent episodes in chronic depression and new-onset cases. For this reason, a question addressing this specificity was introduced in round 2 (n=21), with 62% (13/21) supporting the distinction. For recurrent episodes, most experts favored a minimum duration of 2 weeks (9/13, 69%), whereas for a new onset, opinions varied, with “4 weeks” (5/13, 38%) being the most cited.

In round 2 (n=21), 62% (13/21) emphasized the importance of considering the number of signs or symptoms during screening. Of the 12 experts who specified a threshold, 7 (58%) favored a minimum of 3 symptoms.

Regarding sign or symptom frequency, all experts who provided an opinion in round 2 (20/20, 100%) considered it relevant. Of these 20 experts, 19 (95%) specified a particular frequency, of whom 9 (47%) suggested a frequency of several times per day, 3 (16%) suggested a frequency of nearly daily, and 6 (32%) suggested a frequency of several times per week.

For severity, 90% (18/20) considered it an important factor. Moreover, on a severity scale from 0 to 10 (with 10 being the most severe), the average threshold for clinical concern was 4.9 (SD 1.4), with a median of 5.0 (IQR 2.0).

Further details regarding the results described previously can be found in Multimedia Appendix 5.

The steering committee conducted 2 additional rounds of discussion to decide on the relevance of the 4 dimensions and 18 criteria that did not reach a consensus of ≥70% among the panelists during the 3 e-Delphi rounds.

In the first round of discussion, committee members reviewed the panelists’ comments and agreement levels for each dimension and criterion, analyzing potential reasons for the lack of consensus and considering the clinical implications. Each member shared their perspectives drawing on their clinical expertise. In the second round, the committee revisited the dimensions and criteria after reflecting on the insights from the first round, aiming to resolve any remaining uncertainties and reach a collective decision. Consensus was prioritized, with majority voting used when necessary.

A summary of the committee’s final determinations is provided in Multimedia Appendix 6, where the rationale behind the inclusion or exclusion of each dimension and criterion is detailed.

Table 4 presents an overview of the consensus levels reached in each e-Delphi round and the final decisions made by the steering committee.

In addition to assessing the relevance of each criterion, the panel provided opinions on the preferred assessment method (self-assessment, hetero-assessment, or both). Rather than seeking strict consensus, the goal was to identify general trends. Using a 50% threshold as an example, 35% (28/80) of the criteria were deemed suitable for self-assessment only, 4% (3/80) of the criteria were deemed suitable for hetero-assessment only, and 40% (32/80) of the criteria were deemed suitable for both methods. Detailed information is available in Multimedia Appendix 7, whereas Figure S3 in Multimedia Appendix 3 provides an overview of the consensus rates for these assessment methods.

This modified e-Delphi study aimed to identify and achieve consensus on the most representative signs and symptoms of depression in very old age, focusing on near-centenarians and centenarians.

Our results revealed that most of the items proposed—86% (24/28) of the dimensions and 80% (70/88) of the criteria—which had been assembled based on a literature review, were considered relevant for depression screening in very old age. This highlights the importance of a multidimensional approach and acknowledges the complexity of accurately diagnosing depression in this population.

Focusing on items with the highest consensus rates (≥90%), some overlap between dimensions and criteria was observed, which was expected as dimensions grouped related criteria.

For example, the depressed mood dimension was widely considered relevant, aligning with criteria such as depressed; sadness; and despondency, gloom, and despair, which reflect the emotional core of depressive disorders [6,54]. Similarly, loss of interest in activities, loss of pleasure in activities, lack of reactivity to pleasant events or circumstances, and anhedonia also achieved broad consensus as critical indicators of depression [6,54,62-64]. Collectively, these signs and symptoms highlight the significant impact of depression on a person’s capacity for joy and engagement with life, underscoring their relevance in depression screening.

Criteria such as ruminations and feelings of worthlessness equally achieved a strong consensus due to their role in the negative cognitive and emotional patterns of depression. Previous research has shown that ruminations exacerbate psychopathology, including depression, by intensifying and extending negative mood states, thus interfering with problem-solving and perpetuating physiological stress responses [65]. Similarly, feelings of worthlessness are strongly linked to self-blame, lack of hope, and an increased suicide risk even after depressive episodes remit [66,67]. Building on this, the lack of hope dimension achieved unanimous consensus (21/21, 100%) as a relevant element, reflecting its importance as a significant predictor of psychological well-being [68]. Within this broader dimension, the hopelessness and hopeful about the future criteria also showed high consensus (24/27, 89% and 16/21, 76%, respectively), emphasizing their important role in depression screening in individuals of very old age whether as a negative indicator or a positive resource.

Furthermore, previous episodes of depression or diagnosed depression were also identified as relevant, highlighting the importance of considering an individual’s psychiatric history as past experiences with depression increase the risk of recurrence [69]. In addition, critical life events perceived as negative were recognized for their significant association with higher levels of depressive symptoms in a recent meta-analysis [70].

Finally, the strong agreement on severe criteria such as recurrent thoughts of death or suicide, suicidal ideation, and past suicide attempt(s) underscores the critical need to address these factors due to their immediate risk to individuals’ safety [26].

Having established the most consensual depressive features, it is essential to explore how these characteristics manifest specifically in older adults as unique age-related factors may shape their expression and impact.

As individuals age, they may face different challenges, some of which are particularly relevant to LLD. These include agism; loss of income; the death of loved ones; and the risk of isolation, functional limitations, multimorbidity, cognitive decline, and institutionalization [71]. Such multidimensional challenges can impact well-being and mental health, contributing to reduced aging satisfaction and increased loneliness, including near-centenarians and centenarians [71,72]. While the cumulative effect of these difficulties can increase the risk of LLD, it is important to note that experiencing them does not necessarily lead to the development of a depressive state [72].

The dimensions and criteria that achieved high consensus, as previously discussed, clearly represent central elements in depression screening across all age groups, aligning with current diagnostic standards [6,54]. Our findings confirm their importance for accurate diagnosis in individuals of very old age, underscoring their transversality across generations and relevance even in the final stages of life. While depression in very old age may have unique aspects, some of the core signs and symptoms of the disorder remain consistent and should not be overlooked, including in near-centenarians and centenarians. However, these signs and symptoms may carry additional significance due to the unique psychosocial and physical challenges faced by these individuals [72].

For instance, loss of interest in activities, loss of pleasure in activities, lack of reactivity to pleasant events or circumstances, and anhedonia may highlight a deeper disengagement from life that can be difficult to detect in older adults. These symptoms are often masked by typical aging challenges such as physical decline, social isolation, or even cognitive issues, complexifying their identification. In addition, feelings of worthlessness and the negative impact of critical life events may be particularly relevant in late-life contexts, where significant life changes and losses are common [67,73]. As highlighted in a meta-synthesis [74], older adults often experience declines in physical and cognitive abilities, leading to feelings of disconnection and a perception of being unneeded by their communities. This sense of being unneeded can foster feelings of isolation and insignificance, thereby triggering or exacerbating depressive symptoms. Moreover, societal changes and loss of social roles further deepen these feelings of worthlessness, underscoring the importance of addressing these issues in older populations [74].

While the high consensus on many classic criteria suggests their continued relevance across the life cycle, our findings also point to unique aspects of depression in very old age that may not be fully captured by current approaches. For example, the unanimous agreement on the relevance of lack of hope highlights its particular significance in this age group. Unlike other factors typically associated with normal aging processes, low levels of hope have been identified as a significant risk factor in older adults, reflecting a deeper sense of despair with potentially serious consequences. A systematic review [75] underscored that lack of hope is closely linked to negative outcomes in older populations, including an increased risk of depression and poorer overall health outcomes.

Moreover, the strong agreement on severe symptoms such as recurrent thoughts of death or suicide, suicidal ideation, and past suicide attempt(s) underscores the importance of also addressing these factors in very old age, where such topics often remain taboo [76]. Reluctance to discuss or acknowledge these symptoms may lead to underdiagnosis and inadequate treatment, emphasizing the need for sensitive and proactive approaches in this age group. Recent qualitative research [76] has revealed that suicidal thoughts in later life are often subtly expressed and normalized in everyday conversations, making them difficult to detect. However, the extent to which this applies to centenarians is unclear. Previous research suggests that only a minority of centenarians express a wish to die [77,78], making explicit suicidal thoughts likely rare and not always linked to depressive states. Nonetheless, feelings of life being completed and no longer worth living reflect a deeper disconnection from life, characterized by aching loneliness, feelings of not mattering, and a physical and mental tiredness [79]. Creating safe spaces for informal—yet meaningful—discussions is essential, alongside strategies for timely recognition and management of these symptoms [76].

The increased appetite dimension may have been considered not relevant as it can result from situational factors such as attending a day center or moving to a nursing home, where tailored meals and social dining often enhance the willingness to eat, contributing to an increased appetite. Similarly, the significant unintentional weight gain criterion was likely deemed not relevant for related reasons as previous research has found no association between weight gain and incident depressive symptoms in older adults [80].

Certain cognitive dimensions such as memory problems and language/speech (eg, poor speech) were also classified as not relevant, likely due to their overlap with nondepressive phenomena commonly observed in older adults. This highlights the need for tailored screening protocols that can effectively differentiate between depression and cognitive impairment, conditions that often coexist in individuals of very old age and require distinct therapeutic strategies [81,82]. The criterion the mind is as clear as it used to be was likely deemed not relevant for the same reasons.

Criteria such as hard to get started on new projects; staying home instead of going out and doing new things; finding life exciting, wonderful, and enjoyable; easily annoyed; and getting bored often may be more closely related to motivational factors, social involvement, social support, or underlying health conditions rather than directly indicating a depressive state [83,84]. Similarly, criteria such as thinks most people are better off than him/her, feeling that people are unfriendly or dislike him/her, and feels as good as other people may reflect personality traits, which are relatively stable over the life span and may not necessarily reflect a change in mental health status [85].

Symptoms such as psychomotor agitation illustrate the challenge in differentiating between depression-related symptoms and those arising from other geriatric conditions such as dementia or delirium [86], suggesting that certain depressive features may be less applicable or not sufficient for individuals of very old age [87]. In addition, the full of energy criterion, typically seen as a positive indicator, was not deemed relevant, possibly because it is rare for a near-centenarian or centenarian without depression to report feeling full of energy. This likely reflects their overall physical condition rather than a depressive state, and the absence of this criterion may often be confounded with physical exhaustion linked to frailty [88].

The fear of dying criterion was also classified as not relevant. Although it might be assumed that individuals in the final stage of life would fear death, previous research suggests that death anxiety is generally low among older adults, who are more often concerned with the dying process rather than death itself [89]. Similarly, sufficient financial resources was deemed not relevant. While financial difficulties can impact mental health, their link to psychological distress in older adults appears to be more influenced by underlying factors such as mastery and self-esteem rather than acting as a direct indicator of depression [90].

Research on near-centenarians and centenarians has predominantly focused on physical, cognitive, and social health, with less attention paid to psychiatric aspects [13]. Studies on depression in this population often rely on overall assessment scores from standardized instruments rather than detailed analysis of individual items [72,87,91-94]. This gap, alongside the variety of existing tools for screening for depression in very old age, also motivated this study.

Notably, one study conducted a detailed item-level analysis using a 14-item version of the Geriatric Depression Scale in Portuguese centenarians [88] and found that 51.4% reported feelings of worthlessness, supporting the relevance of this specific criterion in our e-Delphi study. In addition, only 12.3% reported having more memory problems, 43.3% expressed a preference for staying home instead of going out and doing new things, 39.8% reported getting bored often, and 33.3% agreed with the statement thinks most people are better off than they are [88]. Globally, these findings align with our results, where these criteria were considered not relevant in this age group.

While these findings suggest reduced presence of symptoms established for younger populations, it remains unclear whether depression in near-centenarians and centenarians significantly differs from that in other older individuals, such as octogenarians; the key features of depression may not vary meaningfully. Therefore, it is worth questioning whether the development of a tool specifically designed for very old age, including the near-centenarian and centenarian population, is truly necessary. Nevertheless, psychological resources such as resilience appear to be particularly strong in near-centenarians and centenarians [77,95]. These protective characteristics may lead traditional depression screening tools to overlook subtle signs of depression that are masked by these strengths. A tool specifically designed for this age group could better capture the subtle differences and specific needs, enhancing the accuracy and effectiveness of depression screening.

Our findings emphasize the importance of considering the duration, number, frequency, and severity of depressive signs and symptoms in near-centenarians and centenarians. Notably, experts differentiated between chronic and new-onset depression, advocating for a nuanced approach that may significantly impact clinical assessments and interventions. This aligns with existing literature documenting distinct clinical profiles for persistent and nonchronic depressive disorders in adults aged 18 to 79 years [96,97], with persistent cases often involving earlier onset, more severe symptoms, and greater treatment resistance [96]. While these findings are based on younger individuals, they reinforce the necessity of a personalized life span approach that includes rigorous screening and comprehensive clinical strategies for the oldest age groups.

Furthermore, experts in our study highlighted the potential value of combining self- and hetero-assessment techniques to screen for depressive signs and symptoms in very old age. However, this approach may require further consideration. Cognitive limitations at this age may affect the reliability of self-reports, complicating the integration with caregiver observations. Developing strategies to reconcile these differing perspectives—such as guidelines for interpreting discrepancies—could ensure that assessments are accurate and sensitive to the unique challenges faced by individuals of very old age. Flexible, combined approaches remain relevant as research suggests that they enhance diagnostic accuracy and patient engagement [98].

Given the findings of our study, a more comprehensive approach to depression screening in near-centenarians and centenarians may be advantageous. The high consensus among experts regarding certain depressive features highlights their relevance and significance for this particular age group. Importantly, the diagnosis of LLD in adults of very old age can be particularly complex due to overlapping symptoms with other mental or somatic conditions, as previously outlined. This overlap necessitates a differential diagnostic approach that integrates a broader perspective. This emphasizes the need for screening tools that are both comprehensive and sensitive to the unique experiences and needs of individuals of very old age. Implementing such tailored screening methods could contribute to improved mental health outcomes within this population.

Our work extends existing literature by providing a comprehensive overview of a wider range of depressive features, including considerations of their duration, number, frequency, and severity—factors that are less considered in standard depression screenings. This highlights the complexity of diagnosing depression in adults of very old age and confirms the necessity of incorporating these specific aspects into the screening process.

While the e-Delphi method was valuable for identifying key elements in detecting depressive signs and symptoms in those of very old age, a separate validation study is necessary to confirm whether the extended screening process would be more effective compared to current tools. Nonetheless, this study hopefully represents a meaningful step forward in improving the detection of depressive signs and symptoms in this population.

We acknowledge several limitations. First, despite the international and professional diversity of the expert panel, it may not have captured the full spectrum of clinical practices or cultural perceptions related to aging and depression. The relative overrepresentation of nurses among the panelists in rounds 1 and 2 may have introduced a bias toward nursing-specific priorities, such as a stronger focus on the practical applicability of screening criteria in daily care. However, this also reflects their pivotal role in direct care for individuals of very old age, making their insights particularly relevant for developing pragmatic screening criteria. Given this, a groupthink effect in which the dominant profession influenced the consensus cannot be excluded. In addition, our panel lacked experts from Asian countries despite notable recruitment efforts even though these regions have some of the highest proportions of centenarians worldwide. This absence may have introduced a bias toward Western perspectives. At the same time, the concept of depression in very old age may be less clearly defined in some Asian cultures due to cultural background. Second, the selection of a 70% consensus threshold may have influenced the interpretation of the results as it reflects a compromise between inclusivity and stringency, subtly shaping the scope of the consensus achieved. Third, the structured nature of the e-Delphi process may have limited the exploration of nuanced symptoms critical to understanding depression in individuals of very old age, a population whose cognitive decline may complicate symptom presentation. Fourth, some overlap between dimensions and criteria was observed, which may be perceived as a limitation to some extent. However, the steering committee was aware of this overlap and intentionally decided to retain it to confirm certain trends in consensus on specific items.

Despite these challenges, this study’s insights have the potential to stimulate discussions on improving mental health strategies for older populations, especially those of very old age. The diverse and interdisciplinary composition of the panel further contributes to the generalizability of the findings, offering valuable insights to health care professionals and researchers worldwide.

Future research should aim to develop and validate a screening tool specifically designed for depressive signs and symptoms in very old age, including near-centenarians and centenarians. A validation study comparing the extended method with established tools or a gold standard diagnostic reference is crucial to determining its efficacy. In addition, it is important to assess the tool’s effectiveness across different cultural contexts to ensure its broad applicability. The incorporation of Patient-Reported Experience Measures in future developments would also be beneficial in capturing the lived experiences and perceptions of this age group.

This expert consensus could significantly contribute to enhancing early detection and intervention for depression in individuals of very old age, notably by informing the creation of a new screening tool. The findings have the potential to improve diagnostic accuracy and, consequently, allow for the personalization of treatment plans tailored to the unique needs of individuals of very old age.

In conclusion, this study used expert consensus to establish a foundational step toward enhancing the detection and treatment of depression among individuals of very old age, including near-centenarians and centenarians. This study emphasizes the need for cautious, multidimensional, and multidisciplinary approaches to depression screening in this age group, advocating for the continuous refinement of screening tools to address the complex nature of very old age effectively.