Introduction
The concept of “endophenotype” was coined in 1966 by Bernard John and Kenneth R. Lewis in a study of the chromosomal and geographical variability of grasshoppers [Reference John and Lewis1]. Later, this term has been used in multiple fields of medicine to clarify the etiology and pathophysiology of various clinical conditions. The endophenotype, also called the intermediate phenotype, has been used in many ways, mostly to refer to a phenotype that is closer to the biological etiology of the disorders than to the signs or symptoms affected by one or more genes associated with the disease [Reference Gottesman and Gould2]. Specifically, in psychiatry, endophenotypes acquired special relevance when they began to be associated with cognitive functioning [Reference Gottesman and Gould2] and were used to help understand the genomics of schizophrenia and other mental illnesses (MIs), becoming an excellent potential tool for multiple studies in neurobiology, neuropsychiatry, neuropsychology, and heritability [Reference Braff and Tamminga3–Reference Toniolo, Fernandes, Silva, da Silva Dias and Lafer5]. An increasing number of studies have included this concept as the basis for their research, using neurocognitive evaluations instead of genetic and brain morphology tests to identify cognitive endophenotypes [Reference Kim, Kim, Koo, Yun and Won6–Reference Van Rheenen, Lewandowski, Lipschitz and Burdick10].
Several criteria have been proposed to establish that a specific characteristic can be considered an endophenotype of pathology. For instance, it must appear concomitantly with the pathology to be studied; that is, it can be considered as an element of the disease in question, although it is not necessarily a requirement of it but has a high probability of manifesting itself [Reference Faraone, Taylor and Tsuang11, Reference Freedman, Adler and Leonard12]. Another criterion that should be taken into account for the determination of an endophenotype is that it should be “measurable” and “temporarily stable”; that is, it should be more of a “trait marker” than a “state marker” of the disease [Reference Doyle, Wozniak, Wilens, Henin, Seidman and Petty13, Reference Vieta14]. Moreover, an endophenotype must be observable in subsequent measurements, thus introducing a longitudinal perspective in the search for and verification of endophenotypes [Reference Vieta14, Reference Glahn, Knowles, McKay, Sprooten, Raventós and Blangero15]. Furthermore, the presence of similar deficits in the unaffected biological relatives of these individuals favors the use of a genetic substrate for them [Reference Balanzá-Martínez, Rubio, Selva-Vera, Martinez-Aran, Sanchez-Moreno and Salazar-Fraile16–Reference Samamé, Martino and Strejilevich20].
The most commonly used criteria to identify endophenotypes in recent clinical studies are as follows: The phenotype (a) is associated with illness in the population, (b) is heritable, (c) is state-independent (manifests in an individual whether or not the illness is active) but age-normed and might need to be elicited by a challenge, (d) co-segregates with the illness within families, and (e) one of which identified in the proband is found in the unaffected relatives at a higher rate than in the general population [Reference Hasler, Drevets, Gould, Gottesman and Manji21].
In the case of neurocognitive endophenotypes, various studies using several of these criteria have shown that there are neurocognitive deficits in the broad domains of attention, memory, and executive functions in patients with MIs, such as bipolar disorder (BD) or schizophrenia, as well as in their first-degree relatives (FDRs), although the latter in an attenuated manner [Reference Arts, Jabben, Krabbendam and van Os22–Reference Volkert, Schiele, Kazmaier, Glaser, Zierhut and Kopf26]. Similarly, complex polygenicity with a predominance of the genetic component has been established by studies including twins and FDR, facilitating the search for cognitive endophenotypes linked to MIs. However, the outright neurocognitive endophenotypes associated with MIs remain unclear. Moreover, a consensual and standardized cognitive evaluation and selection procedure that allows identification and a more comprehensive description of the cognitive endophenotypes associated with MIs are not available yet [Reference Bourne, Aydemir, Balanzá‐Martínez, Bora, Brissos and Cavanagh27–Reference Maekawa, Katsuki, Kishimoto, Onitsuka, Ogata and Yamasaki29].
Studies on endophenotypes constitute a cost-effective and easier method to implement when studying the wide range of subclinical characteristics of MIs [Reference Balanzá-Martínez, Rubio, Selva-Vera, Martinez-Aran, Sanchez-Moreno and Salazar-Fraile16, Reference Balanza-Martinez, Crespo-Facorro, Gonzalez-Pinto and Vieta23, Reference Cardoso, Bauer, Meyer, Kapczinski and Soares30–Reference Tabarés-Seisdedos, Balanzá-Martínez, Sánchez-Moreno, Martinez-Aran, Salazar-Fraile and Selva-Vera32]. Characterizing endophenotypic profiles associated with MIs could be useful for identifying individuals at risk, increasing the effectiveness of early diagnosis, improving disease subtyping, and proposing therapeutic strategies to reduce the negative effects of the symptoms, in addition to serving as a scientific basis for the physiopathology of the disease [Reference Mora, Portella, Forcada, Vieta and Mur33–Reference Torrent, Martinez-Arán, del Mar Bonnin, Reinares, Daban and Solé35]. Thus, the identification of suitable cognitive endophenotypes for MIs is a potentially useful strategy to improve the understanding of MIs [Reference Dias, Balanzá‐Martinez, Soeiro‐de‐Souza, Moreno, Figueira and Machado‐Vieira36–Reference Russo, Mahon, Shanahan, Ramjas, Solon and Braga39].
Due to the discrepancies found in the procedures and criteria used to identify the cognitive endophenotypes of a MI and in the results on what can be considered a stable cognitive endophenotype of a certain MI, we consider that there is a need for a standardized method that provides definite and clear neuroscientific support for cognitive endophenotypic profiling in future research. The main objective of this study was to present a method that includes and refines the procedures used by other researchers for the search and identification of suitable cognitive endophenotypes in MI research, for each individual diagnosis or for identification of common endophenotypes across multiple diagnoses. We propose an inventory of exploration and verification, which in addition to fulfilling its primary objective, could be useful in unifying the results of previous studies in future investigations.
Methods
Scoping review
We conducted a scoping review (PRISMA-ScR) [Reference Tricco, Lillie, Zarin, O’Brien, Colquhoun and Levac40], which is the most widely used method for synthesizing research evidence when the subject has not yet been extensively reviewed or is complex or heterogeneous in nature. The method is mostly used when researchers seek, among other things, to identify research gaps in the existing literature and attempt to develop a methodological framework for rigorously and transparently mapping the area being investigated [Reference Pham, Rajić, Greig, Sargeant, Papadopoulos and McEwen41], as in our case. The scoping review protocol was accomplished by the members of the research team CB/07/09/0021 of the Center for Biomedical Research in Mental Health Network (CIBERSAM).
The following search string was used on the “Scopus,” “Web of Science,” and “PubMed/Medline” databases: (endophenotype OR intermediate phenotype) AND (mental disorder OR MI OR psychiatric disorder OR psychiatric illness) AND (cognitive OR neurocognitive) AND (first degree OR relatives). The following main filters were applied at convenience: Full text, article records from inception to July 31, 2022, English, and Humans. The inclusion criteria were original articles that focused on the identification of cognitive or neurocognitive endophenotypes of MIs that include any of the following aspects: (a) Studies on patients with a psychiatric disorder and healthy controls; (b) Studies on patients with a psychiatric disorder and relatives of patients; (c) Studies on patients with a psychiatric disorder, relatives of patients, and healthy controls; (d) Studies on the relatives of patients with a psychiatric disorder and healthy controls; and (e) Studies on patients with different types of psychiatric diseases compared with each other. The exclusion criteria were: (a) Studies focused on different types of endophenotypes, such as genetic, physiological, neurological, brain structure, or other health aspects; (b) Studies where only one or several cognitive functions were evaluated in psychiatric patients, without considering the endophenotypic aspect; and (c) Studies not relevant to the study objective.
Method design
A working team of 14 experts from different Spanish research groups was established to design the cognitive endophenotype identification method in four steps. First, a criteria list was established based on the most used criteria found through scoping review. Second, each criterion was defined based on its background. Third, each indicator was established; that is, the manifest properties by which each criterion can be directly identified and measured. Fourth, based on the know-how of experts, group decisions were made considering the importance of each criterion and indicator. The weight or value that each would have, and the method to rate them and verify the endophenotype were considered to set up a scoring system to obtain numerical data. The scoring system can be used to statistically analyze the results in future studies.
To provide content and construct validity, each expert separately evaluated the relevance, coherence, sufficiency, clarity, and weight, of each element of the method based on: general procedure, aspects to consider during the process, criterion, definition, indicator, score, and verification of the criteria, verification of the endophenotype, and the decision tree. Inter-rater reliability was used to score this process.
Results
Scoping review
Following the search string, a total of 5,176 papers were retrieved from the databases (2,114 from PubMed/Medline, 1,763 from Scopus, and 1,299 from Web of Science) as potential papers for inclusion in the study. After applying the filters, removing duplicates, and unifying and refining the searches of each reviewer, 2,620 articles were excluded. The results of the selected 2,556 publications were screened and evaluated, and were further refined based on whether they described the use of a methodology to identify cognitive endophenotypes of psychiatric diseases. Finally, 83 papers were included in this study (Figure 1).
Figure 1. PRISMA-ScR flow diagram [Reference Page, McKenzie, Bossuyt, Boutron, Hoffmann and Mulrow42].
Analyzing the selection criteria of neurocognitive endophenotypes used in the studies, it was found that, the diseases included attention deficit hyperactivity disorder (k = 11), anorexia nervosa (k = 3), autism spectrum disorder (k = 2), BD (k = 21), eating disorders (k = 3), mood disorders (k = 6), obsessive–compulsive disorder (k = 9), substance use disorder (k = 2), schizophrenia, and psychosis spectrum (k = 30), and others. Unaffected relatives in comparison with genetically unrelated controls were included in the 83 studies. Eighty-one of the papers compared a group of patients with healthy controls. At least two repeated measures were included in four studies. Age-normed and clinical variables that could affect the performance were considered in statistical analyses of 55 of the studies.
Based on these findings a list of the “most used criteria” and its “indicators” were configured:
-
1. Concomitance or association with the disease, high probability of manifestation, and measurability: comparison with other groups.
-
2. Presence in biological relatives, heritability, or co-segregation with the disease within families: inclusion of unaffected relatives in comparison with genetically unrelated controls in the study.
-
3. Temporary stability (longitudinal perspective): longitudinal studies with at least two repeated measures.
-
4. State independence: age-normed and clinical variables that could affect performance considered in statistical analyses.
The 83 articles summarizing the target groups, endophenotypes studied, and selection criteria used, are arranged in chronological order in Table 1.
Table 1. Summary of the included articles.
Abbreviations: ADHD, attention deficit hyperactivity disorder; AN, anorexia nervosa; ASD, autism spectrum disorder; BD, bipolar disorder; ED, eating disorders; MD, mood disorders; OCD, obsessive–compulsive disorder; SUD, substance use disorder; SZ, schizophrenia and psychosis spectrum.
a Most used criteria: 1 = Concomitance or association with the disease, high probability of manifesting itself, and measurability; 2 = Presence in biological relatives, heritability, or co-segregation with the disease within families; 3 = Temporary stability (longitudinal perspective); 4 = State independence.
In summary, most of the articles reviewed and evidenced the existence of endophenotypes in individuals diagnosed with MIs and their FDR compared with healthy controls, fulfilling the first and second criteria. Very few of them included repeated measures regarding the third criterion. As to the fourth criterion, although in most of the articles some sociodemographic and clinical variables were controlled, they did not control the effect of some of them that might have given rise to interpretation biases and reduced the power of the findings. Additionally, none of them considered these a useful criterion for the selection of endophenotypes. Another important aspect is that very few studies corroborated their findings on a certain endophenotype based on the same description of a specific cognitive function and/or with the same measurement instrument or even the same clinical type as in other studies.
The designed method
The final version of the method was obtained after refining each of the four steps through a cross-review among the experts. The consensus included five classification categories or criteria: (a) association, (b) heritability, (c) stability, (d) independence, and (e) reliability of results with corresponding descriptions, each with indicators and corresponding weightage and a particular verification system. Although these five categories include the previously used criteria, some modifications were made in terms of their definition and indicators that identify them. For example, the “heritability” criterion synthesizes the following: the endophenotype is heritable, co-segregates within families, and can be seen in unaffected relatives. This criterion also includes a new indicator—the profile or an intermediate pattern of family members’ performance compared with individuals with particular MI and healthy controls [Reference Kim, Kim, Koo, Yun and Won6, Reference Frantom, Allen and Cross67, Reference Islam, Habtewold, van Es, Quee, van den Heuvel and Alizadeh124]. We also added a new criterion, “reliability of results,” referring to corroboration of the findings with those of previous studies because although these criteria are generally reflected in studies of this type, they are not specifically referred to as selection criteria.
Finally, the method with five deliveries was generated: (a) a general procedure and aspects to consider; (b) a reclassification of five criteria, each with its own definition, indicators, and scoring system; (c) verification of the criteria; (d) verification of the endophenotype (Table 2); and (e) a decision tree (Figure 2).
Table 2. MICEmi scoring template.
Figure 2. Decision tree.
Discussion
In recent years, the interest in cognitive endophenotypes related to MI has increased, causing a change from the previous paradigm—which considers the alterations in MIs as irreversible and that treatments are exclusively curative—toward a new paradigm that focuses on the prevention of those impairments. However, its application is hampered by the lack of a consensus, standardized cognitive evaluation, and selection procedure that allows the identification and a more comprehensive description of the cognitive factors associated with MIs. The use of a method to verify specific criteria to study cognitive endophenotypes in a population with MIs can provide some valuable advantages for researchers, such as systematization, replicability, convergence between different clinical findings, and the delimitation of cognitive endophenotypes for each disease.
The proposed method in this study offers a systematic way of identifying and replicating endophenotypes and therefore should be interpreted as a starting point where the primary goal is the exchange of points of view and subsequent contributions to enrich this field of knowledge and to approach the complexity of reality in a more structured way.
In future research, in addition to the criteria that have already been used to identify cognitive endophenotypes in MIs, it is necessary to add other aspects to the analyses that have not always been studied for further improvement. First, to avoid possible misinterpretations of what is being measured, the same test should be performed to measure each cognitive function, or the results of different tests should be comparable in the most valid and reliable way possible. Second, a greater number of repeated measurements should be made with intermediate time intervals, not so close that they generate a training or learning effect but not so distant that they cause a significant decrease in the sample number. Third, whenever possible, three study groups should be included, including patients with MI, relatives, and controls. Fourth, the maximum number of sociodemographic, psychosocial, clinical, and biological factors should be included to rule out any other possible influences on the cognitive function evaluated other than the biological and genetic factors themselves. Lastly, as we propose in our method as the fifth criterion, “reliability of results,” the findings should be corroborated by previous studies.
This work is ongoing, because it is necessary to obtain external validation of the applicability of the method in future research.
Author Contribution
P C-G: had the original idea; conception and design of the study; co-direct the work team; co-coordinated the scoping review; acquisition and analysis of data for the scoping review; drafting the manuscript and figures. JV S-O: conception and design of the study; co-coordinated the scoping review; acquisition and analysis of data for the scoping review; drafting the manuscript and figures. R T-S; V B-M: co-direct the work team; drafting the manuscript and figures. I F-D, A M-A, JC R-R, G S-V, J V-F, D M-SG, C S-M, R A-A: work team member; drafting the manuscript and figures. L R-B, P C-E: acquisition and analysis of data for the scoping review. All authors have read and agreed to the published version of the manuscript.
Data Availability Statement
The data supporting the findings of this study are available from the corresponding authors upon reasonable request.
Acknowledgments
Thanks to all those who have helped in carrying out the research.
Financial Support
This research received no specific grant from any funding agency, commercial or not-for-profit sectors.
Conflicts of Interest
The authors declare none.
Open access
Comments
No Comments have been published for this article.