Thematic research clusters in very old populations (≥ 80 years): a bibliometric approach

We used the National Library of Medicine’s MeSH thesaurus to identify documents (articles, reviews, and letters) that had been assigned the descriptor “Aged, 80 and over” (“a person 80 years of age and older”) between 2000 and 2019. The results comprised two groups of documents:

Searches were undertaken using the Web of Science (Clarivate Analytics), which includes the MEDLINE database, on 24 June 2020.

All bibliographic data were downloaded in order to review and standardize them prior to quantitative analysis. Multiple entries in some fields (institutional affiliations and MeSH descriptors) were individualized, and information on authors’ countries was extracted based on institutional affiliations, unifying all documents assigned to Scotland, England, Wales, and Northern Ireland as originating in the UK.

The WoS-CC databases include subject categories for both “Geriatrics & Gerontology” (in the case of the Science Citation Index Expanded (SCIE), which “includes the clinical, biochemical, histological, and psychological aspects of aging”) and “Gerontology,” linked to the Social Sciences Citation Index (SSCI) and covering “resources that are concerned with the sociological and psychological issues of aging.” For journals assigned to both categories, and generally for all journals assigned to more than one subject category, we made fractional assignments for the documents, proportional to the number of subject categories pertaining to the journal of publication.

As this was an analysis of available published research, no ethics approval was required. No authors were contacted for further information regarding their publications.

We identified 707,175 documents that had been assigned the “Aged, 80 and over” descriptor in MEDLINE and 22,043 documents referring exclusively to this age group and published in journals included in the WoS-CC. The dataset included 16,383 articles (74.3%), 4456 letters (20.2%), 843 reviews (3.8%), and 361 proceedings papers (1.6%).

The annual evolution in the number of documents published (Fig. 2) shows that while this age group has always been present in biomedical research, the boom in papers focusing exclusively on this population began in the mid-2000s.

The distribution of documents by geographic region (Table 1) shows that European researchers are responsible for the greatest total volume of scientific output (43.7% of the documents), followed by North America (30.5%) and Asia (26%). That said, Asian research output grew notably over the research period in both relative and absolute terms, while a certain decline is apparent in the other two regions. The rest of the regions studied made only nominal contributions (0.5 to 4.4%).

With regard to research participation by country (Table 2), the USA ranks first (27.7% of the documents), followed by Japan (14.1%) and different European countries (UK, Italy, Germany, France, and Spain at 5.9 to 7.9%). Japan, Spain, and China are the countries that have seen the greatest growth in scientific activity in this area, while Australia, Japan, and European countries stand out in terms of the number of publications per million population.

The analysis of disciplines in which the documents were published (Table 3) reveals the considerable dispersion of research interests, with investigators in 25 specialties participating in at least 1% of the documents. In absolute terms, the top categories are Cardiac & Cardiovascular Systems, Medicine, General & Internal, and Surgery. According to the size of the category (n journals), the most prominent topics, apart from Geriatrics & Gerontology and Gerontology, were Dermatology and Cardiac & Cardiovascular Systems, and among the moderately productive journals, Anesthesiology and Emergency Medicine.

The co-word analysis of MeSH descriptors showed four prominent clusters, that bring together 77.5% of the documents, reflecting the main topic areas attracting the most research interest in relation to very old populations.

There is a heterogeneous cluster covering closely interrelated disorders and diseases that have a particularly strong impact on people of very advanced age (cluster A). The concepts are linked to diagnosis and treatment of the diseases as well as social and psychological aspects related to aging. The following topics stand out:

Another cluster with a high density of interlinking terminology revolves around heart diseases and the surgical procedures used to address them (cluster B).

A large set of terms linked to neoplasms, tumors, and carcinomas make up another cluster (cluster C), in which aspects such as pharmacological and surgical treatments, biomarkers, and antineoplastic agents are considered. This cluster also shows a specific line of research into “anti- vascular endothelial growth factor A” as a strategy to limit the angiogenesis associated with the use of drugs like bevacizumab, for treating both different cancers and age-related macular degeneration.

Management of infections associated with “staphylococcal infections,” “prosthesis-related infections,” “urinary tract infections,” “klebsiella infections,” “gram-positive bacterial infections,” and “endocarditis, bacterial” are the main research topics in this cluster (cluster D), which covers topics like the combination of drugs, the sensitivity of microbial tests, and bacterial resistance.

Table 4 presents the scientific output and citation degree linked to all of the subject areas described. Overall, geriatric diseases and particularly Health services for the aged & social and psychological issues of aging and Neurocognitive & depressive disorders share the research spotlight with Cardiovascular diseases and Neoplasms.

Additional file 1 presents the list of MeSH terms for diseases assigned to at least 50 of the documents on people aged 80 and over, with indicators for overall output and the number of documents that focus on the topic specifically for this age group. In general, the main diseases either fall under one of the clusters described above or are closely related to them. For example, they may deal with different neoplasms like “breast neoplasms” or “pancreatic neoplasms,” fractures like “spinal fractures,” or neurodegenerative diseases like “Parkinson’s disease.” There are also several diseases that are independently relevant, like “intestinal obstruction,” “deglutition disorders,” and “pulmonary embolism.”

Some descriptors for age-related pathologies are more likely to be associated with descriptors like “aged” (“a person 65 through 79 years of age”) than exclusively “aged, 80 and over,” so they do not appear in the clusters described above. This is the case of “diabetes mellitus,” “osteoporosis,” “pulmonary disease, chronic obstructive,” “depression,” and “Parkinson’s disease.” Some significant differences between men and women have also been observed for certain diseases (Table 5). For example, research into some diseases is more frequently associated with women, such as heart valve diseases (e.g. “aortic valve stenosis”), degenerative diseases of the central nervous system (“dementia” and “Alzheimer’s disease”), falls and fractures (“accidental falls,” “hip fractures”), cardiovascular diseases (“hypertension”), neoplasms (“breast neoplasms”), and other deposition diseases (“calcinosis”). Other diseases are more researched in men, including some neoplasms (“skin neoplasms,” “adenocarcinoma,” “lung neoplasms,” “melanoma,” “prostatic neoplasms,” and “liver neoplasms”) and cardiovascular diseases (“aortic aneurysm abdominal” or “coronary artery disease”).

Our results show a boom in research focused specifically on the oldest population starting in the mid-2000s, with a dramatic rise in the number of publications produced which has continued to the present day. A myriad of factors contributes to explaining the growth in scientific research output. First and foremost is the increased life expectancy and population aging [29]. At the same time, social and political consideration towards older people’s problems and conditions are favorable, with increasing investments in research projects focusing on concepts like healthy aging and well-being in the elderly [6, 8, 9]. Professional training in geriatrics has been developing, too, with the implementation of specialized education programs and the consolidation of this field as an established specialty, different from internal medicine and promoted by its own prestigious scientific societies [13].

The pronounced leadership of European research over that of the USA, whose output levels are below those seen in other biomedical research areas [30] is another notable finding from our analysis. At a country level, the importance of Japan, as the second biggest producer of research on very old populations, is noteworthy. As in Europe, this situation can be explained because Japan has one of the highest rates of population aging and among the longest life expectancies worldwide [31]. Logically then, researchers from these settings have typically occupied prominent positions in quantitative analyses of the scientific output on diseases like stroke, urinary tract infections, and eye diseases, which disproportionately affect older populations [20, 32, 33]. In contrast with other research areas, we did not observe a boom in publications from countries like China [34]. Our results, which point to a concentration of research in Europe and North America as well as in Japan, suggest the need to promote research focusing on very old populations in other settings, like southeast Asia, where the pace of population aging is accelerating [35].

The numerous disciplines linked to scientific output on very old populations confirm Mussi et al.’s [12] conclusions on the multidisciplinary nature of this research area. Indeed, the literature focusing on this collective is widely dispersed among specialized journals in different areas. Geriatrics and gerontology represent horizontal disciplines, treating elderly people who often have different, generalized problems associated with aging or geriatric syndromes. At the same time, old people also have diseases linked to other organs and organ systems that must be addressed by specialists from other disciplines, and this relationship is reflected in the corresponding research streams.

Dementia occupies a preeminent position in the research performed, as this is the main cause of dependence and disability among very old people [36]. In fact, in the UK the associated costs of this group of diseases exceed the combined costs incurred by heart diseases and cancer [37]. Since the turn of the 21st century, then, dementia has emerged as a priority for health systems in high-income countries [38].

Alzheimer’s disease is positioned as the main entity studied in research on age-associated dementias (n = 575), confirming the results reported by Serrano-Pozo et al. [17]. These authors analyzed the literature on Alzheimer’s disease published from 1975 to 2014, observing a steady growth of research output over the study period, especially in publications focused on the 80+ age group starting in the 1990s. Alzheimer’s disease is the most common type of dementia, accounting for up to 70% of all cases [6, 15, 39], and its prevalence and associated mortality has seen the most pronounced growth over the past several decades in the very old population [4]. Nevertheless, the relative dearth of research into other kinds of dementia (“Dementia, Vascular” n = 44; “Lewy Body Disease” n = 44; and negligible research in other pathologies) shows the need to promote research focusing on dementias other than Alzheimer’s in very old people. The scant research identified on “Frontotemporal dementia,” which is the second cause of dementia, is probably due to its earlier onset compared to Alzheimer’s [40].

Frailty constitutes another popular research topic associated with aging, as this is a very prevalent geriatric syndrome (≈ 30%) in people aged 80 years or older. Clinical manifestations include muscle weakness, fatigue, and slow motor performance, among others, and these are often accompanied by social vulnerability and dependence [6] along with a higher risk of falls, depression, disability, and mortality [41]. The ongoing COVID-19 pandemic stands out as the clearest example of how frailty can contribute decisively to situations of vulnerability in very old people and to a more severe clinical presentation of other diseases [42, 43].

The mingled impacts of different clinical and social factors underline why clinical research in very old people must necessarily take into account other social and technological considerations in order to reduce disease incidence and associated harms, both in patients and in their family members and caregivers [16]. Our findings on the relevance of caregiving, specialized care, ethics, and social and care support for patients with dementia and their family members corroborate those reported by Baldwin et al. [24].

More recently, Shi et al. [25] studied research on family caregivers of people with dementia, highlighting that the increased prevalence of dementia in very old people and their lengthening life expectancy can increase the risk of diverse health problems in family members, including depression, stress, anxiety, and other physical or psychological disorders. Caregivers have even been called “silent patients” or “secondary patients” whose needs should be addressed through specific strategies to protect and provide them with the required support; such strategies show unequal development across different countries [16]. For their part, Asghar et al. [44] studied assistive technology for people with dementia, signaling the relevance of telecare devices, activity monitoring, warning and reminder systems, fall detection, ease of mobility and communication, among other related aspects, tied to concepts like “independent living,” or “activities of daily living” in the present study.

The development of research in the area of fracture surgery is also notable, showing substantial growth over the past several years. Hip fractures are the most relevant surgical pathology in people of advanced age, due both to the high incidence and the potential for negative effects on quality of life, autonomy, and level of frailty in this collective. These impacts have driven important advances in areas like minimally invasive surgery and complex fracture management [19].

The descriptors referring to nutrition underscore the relevance of diet in relation with preventive and therapeutic treatments in very old people, as well as for addressing the deterioration of nutrition associated with factors like difficulty swallowing, loss of appetite, and reduced access to healthy diet [18]. Conversely, none of the clusters contain descriptors related with exercise or physical activity (e.g. “exercise” n = 121 or “exercise therapy” n = 57). Efforts to promote the integration of investigators from these areas into research on the pathologies identified could be warranted. Physical activity in relation to aging constitutes an area of active research, as Müller et al. [7] have pointed out. These authors report that the most highly cited literature in this area studies the association between physical activity and health-related outcomes, which may be a relevant factor for preventing dementia. Moreover, once the benefits of physical activity have been established, this study calls for continued research into designing effective and economically feasible interventions for physical activity. For their part, Gu et al. [6] draw attention to different factors in addition to physical activity, like emerging trends of research associated with healthy aging, such as diet, working memory, and active aging, pointing to the relevance that anti-aging medicine may have as a novel research field that considers all the above-mentioned aspects as well as genetic factors associated with aging and anti-aging pharmacological research.

Two prominent research clusters, focused on pathologies that are not specifically associated with very old populations, are devoted to cardiovascular diseases and neoplasms. Coronary artery disease and stroke, in particular, are the main causes of death worldwide and in all regions except Africa, explaining the importance of these areas in research [32].

Other cardiovascular diseases studied in very old patients are aneurysms and myocardial infarction, which follow stroke as the top cardiovascular diseases studied in this population. The rising prevalence of these diseases [45], along with major advances in research, are behind this trend, as described in different bibliometric studies [46‐48]. The prevention and management of these pathologies also receive considerable attention [49], for instance the control of risk factors like hypertension [50] and the application of artificial intelligence as a decision-making tool for the clinical management of stroke and heart disease [51].

Likewise, a substantial body of research revolves around rehabilitation and patients’ adaptation to their new health state and living conditions following a cardiovascular event. Chow et al. [32] points out the importance of this research in driving improvements in patient outcomes. Indeed, the increase in scientific output on stroke contributed to reducing its associated mortality, and in turn, these positive trends have led researchers to direct more attention to aspects like pain management and rehabilitation.

The presence of lung and colorectal cancer in the cluster on neoplasms responds to the fact that these are among the cancers causing the highest mortality in very old people [3, 4]. On the other hand, the importance of skin neoplasms in our results is probably due to different factors, for example, the aggressiveness of melanomas among elderly people or the challenges for oncological practice derived from treatment complexity in people with multiple comorbidities [52]. Other important issues have to do with the fragility of aging skin and environmental factors like solar radiation, making skin care in very old people a priority for dermatological research [53].

The cluster on neoplasms also includes an important line of research on anti-vascular endothelial growth factor A (anti-VEGF) treatments, both for different cancers and for age-related macular degeneration. Although the latter is not a neoplasm, it is the primary cause of vision loss in old people and is mediated by VEGF. In any case, there are still many unanswered questions along this research stream, for example related to the best anti-VEGF medications or the identification of biomarkers that would enable better assessment of treatment efficacy [54].

Finally, the cluster around bacterial infections and anti-bacterial agents shows the importance of clinical entities like urinary tract infections and infections associated with the implantation of prostheses, especially those caused by staphylococci and Klebsiella spp., in very old people [55]. Urinary tract infections—often aggravated by bacterial resistance [56]—are the most common infections in this population due to the degenerative nature of the urinary tract [57], and they are also responsible for worsening patients’ functional status. Research addressing this problem has seen notable growth in the past several years around questions like multidrug resistance and the most appropriate treatments for different types of infections [33].

The main limitation of this analysis is that quantitative studies like ours provide a vision based on the research generating the most publications; however, they do not necessarily capture prominent lines of research on emerging or highly specific pathologies. Our results may also underestimate the body of research on diseases that generally develop before people reach such an advanced age, like diabetes, osteoporosis, chronic obstructive pulmonary disease, or Parkinson’s disease. The individual descriptors for these diseases all stand out as important, but they are not integrated in any of the thematic clusters we identified. Other highly prevalent conditions, for example eye diseases like glaucoma or cataracts, are underrepresented in our findings because there are already well-established pharmacological or surgical treatments available for them, so they attract less research interest. Our use of a descriptor for a human age group led to the exclusion of a large volume of literature that does not exclusively focus on the oldest age groups and probably most basic research that is not explicitly linked to studying very advanced age as a key variable. Finally, our analysis may not have picked up on the full body of research in some psychosocial areas, as it was based on terminology from the MeSH thesaurus, which is mainly rooted in biomedical fields.

In addition, although we observed some very well-defined thematic clusters, multidisciplinary approaches to different pathologies remain vitally important in this population. Not only do a large proportion of very old people have multiple comorbidities, geriatric syndromes like sarcopenia or other conditions provoking the loss of skeletal mass can act as risk factors for pathologies like cardiovascular diseases. These associations are sometimes underrepresented in research [18].