Usefulness of functional assessment of advanced activities of daily living compared with cognitive assessment, as a predictor of mild cognitive impairment in adults older than 60 years: A systematic review

Abstract

Introduction

Functional status includes a continuum from normal ageing to the onset of dementia. Advanced activities of daily living (AADL) are the first activities to be affected in mild cognitive impairment (MCI). This study aims to assess whether AADLs may be used as a predictor of MCI.

Development

A literature search was conducted on the PubMed, Embase, Scopus, CINAHL, Web of Science, and Cochrane databases. Six studies reported that instruments assessing AADLs presented optimal correlation with cognitive measures and good discrimination between healthy controls (HC) and patients with MCI (P < .01) and concluded that they generally present good psychometric properties.

Other studies correlated items assessing AADLs with cognitive assessment instruments or cognitive functions, eg, associating executive functions with AADLs. Some studies report that functional outcomes may improve the clinical classification of MCI and the absence of cognitive impairment.

Some studies also relate AADLs to cognitive functions, seeking to establish specific syndromic measures of functional changes in clinical progression.

The De Vriendt AADL instrument, Brussels Integrated Activities of Daily Living Inventory, and Instrumental Activities of Daily Living-extended scale cannot be used as independent diagnostic tools for evaluating MCI, but should be included in a comprehensive multidisciplinary assessment.

Conclusions

AADLs may be a sensitive measure for early detection of impaired cognitive function. Functional assessment of AADLs (including psychometric assessment instruments) may detect cognitive differences between the healthy population and individuals with MCI, and may be a predictor of MCI.

Resumen

Introducción

En el estado funcional existiría un continuo desde el envejecimiento normal hasta el inicio de la demencia. Las actividades avanzadas de la vida diaria (AAVD), serían las primeras en afectarse en el deterioro cognitivo leve (DCL). El objetivo es valorar si las AAVD podrían ser un predictor de DCL.

Desarrollo

Búsqueda de estudios en las bases de datos: PubMed, Embase, Scopus, CINAHL, Web of Science y Cochrane. En 6 estudios se observó que, instrumentos de valoración en AAVD presentaron una óptima correlación con las medidas cognitivas y una buena distinción entre controles sanos (CS) y DCL (p < 0,01), concluyendo en general, buenas propiedades psicométricas.

Otros estudios correlacionan ítems que evalúan AAVD con instrumentos de valoración cognitiva o con funciones cognitivas, por ejemplo, entre funciones ejecutivas y AAVD. Se ha observado en alguno de los estudios, que las medidas de resultados funcionales pueden mejorar la estadificación clínica, entre la ausencia de deterioro cognitivo y DCL.

También encontramos estudios que correlacionaron AAVD y funciones cognitivas, para tratar de establecer métricas específicas sindrómicas indicativas de cambio funcional a nivel clínico evolutivo.

Los instrumentos psicométricos a-ADL, BIA y IADL-X, no pueden usarse como herramientas de diagnóstico independientes para evaluar DCL, pero sí que deberían incluirse en una evaluación multidisciplinaria integral.

Conclusiones

Las AAVD pueden ser una medida sensible de detección precoz de disminución en funciones cognitivas. La valoración funcional en AAVD (incluyendo los instrumentos psicométricos de valoración en AAVD) podría detectar diferencias cognitivas entre población sana y con DCL, resultando ser un posible predictor de DCL.

Introduction

Currently, more than 47 million people worldwide live with dementia, a number expected to grow to over 82 million by 2030 and 152 million by 2050.¹ As no curative treatment is yet available for the majority of cases, early diagnosis is fundamental.

Neuropsychological evaluation is crucial in the diagnosis of cognitive impairment and dementia, as assessment is difficult in populations with low levels of education and in multicultural settings where it may be influenced by language and education. Therefore, some authors consider functional assessment to be a more ecological approach in this evaluation.² However, it is sometimes difficult to determine how cognitive changes are associated with changes in everyday functioning, and to what extent variation in functional status may be attributed to cognition³; it is also possible that some cognitive domains, such as executive function, may be more relevant than others in functional impairment.⁴

The term activities of daily living (ADL) refers to all those activities that an individual performs on a daily basis, which enable them to live independently, integrated into their environments and fulfilling a specific social role.⁵ ADLs are classified as basic (BADL), those needed to meet basic needs and self-care⁶; instrumental (IADL), more complex activities that are nonetheless essential to independent living (eg, shopping, cooking, and managing medication)⁷; and advanced (AADL), which the first Spanish reference on these activities⁵ defines as elaborate behaviours to control the physical and social environment, enabling an individual to develop a social role, maintain good mental health, and enjoy excellent quality of life. These activities, also known as complex or extended ADLs, refer to volitional activities related to cultural or motivational factors, going beyond what is strictly necessary for independent living, and include leisure and/or self-development activities, as well as (semi-) professional work.^8,9

Many studies report an association between greater cognitive impairment and poorer functional status.⁹ Thus, different authors have demonstrated a significant relationship between everyday functioning and memory,^10,11 executive function,^12,13 and complex reasoning.¹⁴ However, the strength of the reported association varies between studies. One review¹⁵ exploring the cognitive correlates of functional status found that only a modest part of the variation in everyday functioning may be attributed specifically to cognition. Cognition accounts for a mean of 21% of variance in functional outcomes (median, 15.9%). This finding was confirmed in a sample of patients with mild cognitive impairment (MCI).³ That study found that all cognitive domains were correlated with IADLs, though these correlations were very weak (0.053–0.155). Therefore, it is essential to promptly detect subtle functional limitations and to understand their cognitive and non-cognitive correlates.^12,16,17

Degradation of functional capacities occurs gradually and hierarchically, with involvement of IADLS and AADLS preceding the degradation of BADLs. MCI, the intermediate state between normal cognition and dementia, was long thought not to affect functional status,¹⁸ although this position has been questioned more recently^19,20 and authors have begun to argue that, as in cognitive impairment, functional status includes a continuum from normal ageing to onset of dementia.²¹ Furthermore, it has been suggested that functional difficulties are linked to cognitive impairment and may even precede the cognitive deficits observed in MCI.²²

With regard to this functional continuum, AADLs are particularly relevant, as these would be the first activities affected in a process of cognitive decline. This concept was introduced in 1989 by Reuben et al.,²³ and it has subsequently been suggested that, in the assessment of cognitive disorders, evaluation of BADLs and IADLs should be complemented by evaluation of AADLs in order to detect and diagnose cognitive impairment at early stages.^24,25

Therefore, MCI is currently considered to be associated with mild changes in AADLs and even IADLs,²⁶ with disability in these activities representing a strong predictor of future dementia.¹⁷

The literature includes numerous studies addressing the analysis of more advanced functional abilities and whether impairment of these abilities might reflect a very early stage of cognitive decline.^5,27–32 Nonetheless, this question has not been addressed in any systematic review to date; in the light of this, and given our interest in the subject, we considered it important to conduct this systematic review, which we hope may contribute important evidence on this issue.

In this regard, our research is based on^27–32:

• -

  The idea that AADLs may be a sensitive measure for early detection of decline in cognitive function^28,33 and may, therefore, represent a predictor of a process of MCI, alongside neurological and neuropsychological examination, defined as Global Deterioration Scale (GDS) stage 3, Functional Assessment Staging Test (FAST) stage 3,^34–36 or Clinical Dementia Rating Scale (CDR) score of 0.5.³⁷

• -

  The need to consider functional evaluation of AADLs at a similar level of importance to that of cognitive evaluation,⁵ in the assessment of patients with suspected MCI.

Therefore, we conducted a systematic review to establish whether, like cognitive assessment, functional assessment of AADLs might be a predictor of MCI.

Material and methods

A systematic review of the literature was conducted according to the 2020 PRISMA guidelines. The protocol was registered on the PROSPERO register (CRD42023422049). The search was conducted on the PubMed, Cochrane Library, Embase, Scopus, CINAHL, and Web of Science databases, with the following search strategy: mild cognitive impairment OR mild cognitive decline OR mild cognitive diseases OR mild cognitive disorders AND advanced activities of daily living in multidisciplinary databases.

Selection criteria

We included original and complete studies, following a randomised clinical trial (RCT) or observational design (cross-sectional, cohort, and case–control studies), year of publication (given the scarcity of information on this subject) from the beginning of indexing to May 2023, written in English or Spanish, with participants aged 60 years or older, and measuring AADLs and exploring the role of loss of AADLs as a predictor of cognitive impairment.⁵ This functional assessment was compared against cognitive assessment, measured with neurological and neuropsychological examinations in patients with GDS stage 3, FAST stage 3,^34–36 or CDR scores of 0.5.³⁷ We only included studies published in scientific journals with an impact factor, which explained in sufficient detail the methods and procedures followed, and which used validated, reliable psychometric instruments and neuropsychological measurement tools. We also accounted for whether results were theoretical or practical and useful at a social level, the references cited, and compliance with ethical standards and norms. Studies were screened by reading the titles and abstracts; this was performed independently by 2 authors, with a third reviewer resolving any disagreements. Subsequently, 2 reviewers reviewed the full texts of the articles, with a third reviewer resolving any disagreements. The Rayyan systematic review management platform (https://new.rayyan.ai/) was used to support this process.

Data extraction and study analysis

Data were extracted by all members of the research group. For each study, the following information was extracted: study design, number of participants, age of participants, place where the study was conducted, study objectives, psychometric instruments used for functional assessment and clinical and neuropsychological evaluation tools, diagnostic criteria for MCI, cognitive functions assessed, AADLs assessed, results on the correlation between AADLs and cognitive functions, conclusions, and limitations. A total of 11 studies were ultimately included in the systematic review (Fig. 1).

Figure 1.

Procedure for the selection of articles for review.

Risk of bias assessment

The Newcastle-Ottawa scale (case–control and cohort versions) was used to assess the risk of bias. Prior to this analysis, we established that studies presenting high risk would not be excluded in the light of the limited evidence available on the subject. The risk of bias assessment revealed that, of the 11 studies included in the review, risk of bias was low in 8 studies and high in 3. Methodological quality was acceptable to good in all studies. Given the great heterogeneity between studies, we performed a qualitative/narrative analysis. The results of the analysis are summarised in Table 1.

Table 1.

Risk of bias assessment and methodological quality according to the Newcastle-Ottawa scale.

Author, year	Type of study	Selection	Comparability	Exposure	Conclusion
Peña-Casanova et al.,^38 2005	Case–control study	****	**	**	Low risk
De Vriendt et al.,^8 2013	Case–control study	****	**	**	Low risk
Vermeersch,^39 2015	Case–control study	****	**	**()	Low risk
De Vriendt et al.,^40 2015	Case–control study	****	**	**()	Low risk
Bidzan et al.,^41 2016	Cohort study	()()*()	*	Result***(38% lost to follow-up)	High risk
Cornelis et al.,^43 2018	Case–control study	**()*	**	**()	Low risk
Fieo et al.,^42 2018	Cohort study	****	*	**	Low risk
Cornelis et al.,^44 2019	Case–control study	***	**	**	Low risk
Kalligerou et al.,^45 2020	Retrospective, cross-sectional cohort study	***	*	*	High risk
De Vriendt et al.,^46 2021	Case–control study	***	**	**	Low risk
González et al.,^47 2022	Cohort study	***	*	**	High risk

Results

General characteristics of the studies reviewed

A total of 11 studies (8 cross-sectional and 3 longitudinal)^8,38–47 were definitively selected for inclusion in the systematic review; all studies were published between 2015 and 2022 and directly or indirectly evaluate the utility of functional assessment of AADLs, in comparison to cognitive assessment, as a predictor of MCI in patients aged 60 years or older. All studies were non-randomised (no RCTs on the subject were identified in the literature search), with 4 cohort studies (observational and analytical prospective studies) and 7 were case–control studies (observational and analytical). Table 2 summarises the studies reviewed.

Table 2.

Main characteristics of the studies included in the systematic review.

Author (year), country	N	Sample characteristicsGroupsMen/women	Source of data	Age, mean (SD)(range)	Study objective	Diagnostic criteria	Statistical method/type of study/follow-up period	Psychometric instruments used: clinical and neuropsychological assessment/functional assessment	Cognitive functions and AADLs assessed	Results	Conclusions	Limitations
Peña-Casanova et al.^38 (2005)Spain	107	51 men and 56 women42 HC19 MCI46 AD	NORMACODEM projectHC: Hospital del Mar de Barcelona, Hospital Clínic y Provincial de Barcelona, and Hospital Mútua de TerrassaMCI/AD: Hospital del Mar de Barcelona and Hospital Clínic y Provincial de Barcelona	64.27 (10.14) years	To establish the possible correlations between a-BT global score and functional scales of ADLs to determine the extent to which cognitive impairment is related to or correlated with changes in ADLs.	NINCDS-ADRDA.CDR	Pearson correlation coefficientCase–controlCross-sectional	a-BTRDRS-2BDRSIDDD(GERRI in the NORMACODEM study)	a-BT: Orientation, language, reading, writing, visual recognition,memory, and abstraction.BDRS: AADLs (item “reduced interest in usual activities”).GERRI (NORMACODEM study, from which the subsample used in this study was extracted): AADLs (items 9, 17, 31, 36, and 42).	A good functional correlation was observed with a-BT global scores and AADLs, at least in patients with MCI.All cognitive-functional correlations were significant (P < .0001) and strong, rangingfrom 0.76 to 0.80.	The a-BT enables the prediction of subjects' functional prognosis.	As an explicit limitation, the authors note that below a moderate–severe degree of impairment in AD, a-BT scores present a floor effect (35 points) and do not permit subsequent functional correlations.
De Vriendt et al.^8 (2013)Belgium	68	26 HC17 MCI25 mild AD	GDH of UZ Brussels University Hospital and Ghent University Hospital	80 years (4.6)(66–90)	To test a new psychometric instrument (the a-ADL, developed according to the ICF) for the assessment of AADLs, with a view to confirming whether it discriminates between HCs, MCI, and mild AD.	HC (excluding any objective functional or cognitive deficits suggestive of a diagnosis of MCI or AD [MMSE < 25/30, CAMCOG < 80/105]).MCI (International Working Group on Mild Cognitive Impairment criteria)AD (DSM-IV or NINCDS-ADRDA criteria)	Kolmogorov–Smirnov testIntraclass correlation coefficientKruskal-Wallis testMann–Whitney U testWilcoxon signed rank testChi-square testROC curveCase–controlCross-sectional	MMSECAMCOGGDS-YesavageNPI-QKatz IndexPGC-IADL (modified gender-specific version)Physical examinationInventory of comorbidities and medications usedBlood analysisNeuroimaging (CT or MRI)	MMSE: 1) orientation to time and space; 2) memory registration; 3) attention and calculation; 4) recall; 5) language and constructive praxis.CAMCOG: orientation, language, memory, praxis, attention, abstract thinking, perception, and calculation.a-ADL: total number of activities performed by a person, taking each subject as their own reference.a-ADL-DI, a-ADL-CDI, and a-ADL-PDI.49 AADLs, divided into 15 groups according to the ICF.	a-ADL-CDI differed significantly between all 3 groups (P < .01).a-ADL-DI differed significantly between MCI and AD (P < .001).a-ADL-PDI did not differ significantly between groups.The a-ADL had good psychometric properties (inter-rater reliability; agreement between patient and proxy; correlations with cognitive tests).	The a-ADL-CDI and a-ADL-DI may represent a useful contribution to the identification and follow-up of MCI in older adult populations.The scoring system of this novel tool enables the detection of the subtle changes in functional status that occur in MCI.It is able to discriminate normal age-related decline from that occurring in MCI and AD.Functional assessment of AADLs may constitute an important predictor of progression to AD.	As an explicit limitation, the study only indicates that given their sensitivity and specificity, the a-ADL-DI and a-ADL-CDI should be used with caution.The authors also note that, as MCI represents a heterogeneous condition, a longitudinal study is needed to establish whether these indices are able to predict which patients will remain stable and which will develop dementia.
Vermeersch et al.^39 (2015)Belgium	143	50 HC (19/31)45 MCI (22/23)48 AD (12/36)	HC: call for volunteers or word of mouth in the communityMCI/AD: patients attending a GDH for the first time due to memory complaints	HC: 79.5 years (5.1)(65.9–91)MCI: 80.39 years (4.8)(67–90)AD: 80.8 years (5.1)(69–91)	To explore the relationship between functional impairment in the 3 domains of the a-ADL (developed according to the ICF) and cognitive impairment in HCs and patients with MCI and AD.	International Working Group on Mild Cognitive Impairment criteriaDSM-IVNINCDS-ADRDA	One-way ANOVAPost-hoc Scheffe testChi-square testPearson correlation coefficientCohen conventionsCase–controlCross-sectional	MMSECAMCOGb-ADLi-ADL a-ADL a-ADL-CDIa-ADL-CDI-ETa-ADL-CDI-DRIVINGa-ADL-CDI-ECON	Use of everyday technologyDriving a vehicle (DRIVING)Performing complex economic activities (ECON)	The a-ADL-CDI-ET and cognitive disability index for all 3 AADLS together differed significantly between the 3 groups.a-ADL-CDI was strongly correlated with cognitive measures.	Functional impairment of certain AADLs may be an early marker of cognitive impairment.	The study's cross-sectional design is unable to demonstrate causality.Cognitive assessment was not comprehensive.Potential informer bias.Data from the HC group were self-reported.
De Vriendt et al.^8 (2015)Belgium	150	50 HC48 MCI52 mild–moderate AD	HC: cognitively healthy community-dwelling volunteersMCI/AD: 2 consecutive groups of community-dwelling patients drawn from GDHs	HC: 79.5 years (5.1)(65.9–91)MCI: 80.3 years (4.6)(67–90.4)AD: 81 years (5.2)(69–91)	To evaluate the discriminative validity of the a-ADL (developed according to the ICF) for MCI.	International Working Group on Mild Cognitive Impairment criteriaAD (DSM-IV or NINCDS-ADRDA criteria)	One-way ANOVABonferroni testANCOVAROC curveCase–controlCross-sectional	MMSECAMCOGb-ADLi-ADLa-ADL-DIa-ADL-CDIa-ADL-PDI	49 AADLs divided into 15 groups	The a-ADL can discriminate between study groups with satisfactory validity.a-ADL-DI and a-ADL-CDI differed significantly between groups; a-ADL-PDI did not.	The a-ADL has good capacity and satisfactory validity for distinguishing cognitively normal ageing, MCI, and AD.Follow-up with functional assessment of AADLs may constitute an important predictor of progression to AD.	Lack of clear consensus in the definition and standardised assessment of MCI.Data from the HC group were self-reported.
Bidzan et al.^41 (2016)Poland	193	1 cohortPatients with MCI (MMSE 24–30 or GDS/Reisberg stage 3)	Clinic for Developmental Psychiatry, Psychotic Disorders and Advanced Age Studies, Medical University, Gdańsk	77.3 years (9.18)	To establish the relationship between the level of activity (intellectual, physical, and social) and future progression to cognitive impairment in patients with MCI.	International Working Group on Mild Cognitive Impairment criteria	Two-tailed test (P < .05)Chi-square testTest for 2 variancesCohort studyLongitudinal7 years	MMSEActivity Scale(intellectual, physical, and social subscales)PGC-IADL	MMSE: 1) orientation to time and space; 2) memory registration; 3) attention and calculation; 4) recall; 5) language and constructive praxis.Activity Scale1) Intellectual (5 items)2) Physical (3 items)3) Social (5 items).PGC-IADL: 8 items	Significant changes in global scores on the Activity Scale and IADL scale.Significant changes on the IADL subscales for telephone use, shopping, transportation, and personal finances.The Activity Scale only showed significant changes in the physical subscale.	Assessment of the level of activity may be useful for predicting the future progression of MCI.	The impact of medications on cerebral neurodegenerative mechanisms and their potential impact on cognitive test results (MMSE) appear to be an essential factor.The quantitative evaluation of individuals' level of activity will always be subject to considerable simplification of the whole phenomenon.
Fieo et al.^42 (2018) United States	2471	No cognitive impairment at study start (HC)1853 remained cognitively unimpaired (34%–66%).618 progressed to MCI (29%–71%)	Cohort from the WHICAP project (community sample)	76.2 years (6.3)HC: 76.1 years (6.2)MCI: 76.5 years (6.1)	To examine the predictive power on the IADL-x scale for MCI.	MCI (Petersen criteria)	Cox regressionMultinomial ordinal regressionCohort studyLongitudinal5.1 years (2.4)	The cognitive tests used are not specified.IADL-x scale (5 AADLs + 4 IADLs)	Cognitive functions: not specified.AADLs: IADL-x scale (9 items: 5 AADLS + 4 IADLs).	Subjects scoring below the total median score for 6 activities on the IADL-x scale were 2.5 times more likely to develop MCI at 2 years and 2 times more likely at 5 years.	The IADL-x scale seems to be a valid tool for assessing the risk of progression to MCI.	The scale's dichotomic format (yes/no) reduces the amount of information gathered from each subject.
Cornelis et al.^43 (2018)Belgium	61	21 HC20 MCI20 AD	GDH of UZ Brussels University Hospital	HC: 78 years (10)MCI: 79.5 years (4)AD: 80 years (9)	To evaluate the convergent and concurrent validity of the a-ADL, i-ADL, and NAT tools in order to compare their diagnostic accuracy in discriminating between HCs and patients with MCI and AD.	MCI (Petersen criteria)AD (NINCDS-ADRDA)	Kruskal-Wallis testChi-square testMann–Whitney U testSpearman R correlation coefficientROC curveCase–controlCross-sectional	MMSECAMCOGKatz IndexPGC-IADLNATi-ADLa-ADL	MMSE: 1) orientation to time and space; 2) memory registration; 3) attention and calculation; 4) recall; 5) language and constructive praxis.CAMCOG: orientation, language, memory, praxis, attention, abstract thinking, perception, and calculation.NAT: task 1: making toast and coffee; task 2: wrapping a gift as a present; task 3: preparing and packing a child's lunchbox and packing a child's schoolbag.i-ADL: household activities (9 items).a-ADL: evaluates complex activities.	NAT: tasks 1 and 3 differed significantly between the 3 groups.Task 2 only differed between HC and MCI and between HC and AD.The i-ADL, a-ADL, a-ADL-DI, and a-ADL-CDI differed significantly between groups, whereas the a-ADL-PDI did not.Convergent validity:NAT (total) was strongly correlated with a-ADL-DI, a-ADL-CDI, and i-ADL-CDI (r = 0.634–0.663; P < .01).a-ADL-PDI and i-ADL did not present convergence with the NAT (P > .05).	The i-ADL, using a report-based method, and the NAT, based on performance, are not significantly different for discriminating between HC, MCI, and AD.Both methods present strong concurrent and convergent validity and are valid and reliable assessments of AADLs with similar discriminatory power in the diagnosis of cognitive disorders in older adults.	Small sample size, low statistical power.Only one measurement.Data from HCs were self-reported.Cognitive impairment was not ruled out in proxies in the other 2 groups.
Cornelis et al.^44 (2019)Belgium	120	44 HC41 MCI35 AD	GDH of UZ Brussels University Hospital	80.4 years (5.1)HC: 78.9 years (5.1)MCI: 81.1 years (4.6)AD: 81.6 years (5.3)	To examine the relationship between executive functions and BADLs, IADLs, and AADLs.	HCs: patients attending a GDH for non-cognitive complaints.MCI (Petersen criteria)AD (NINCDS-ADRDA)	ANOVAChi-square testBonferroni testPearson correlation coefficientMultivariate analysis with multiple linear regressionCase–controlCross-sectional	MMSECAMCOGTMT-ATMT-BFABKatz IndexPGC-IADLSemi-structured interview: 6 BADLs, 9 IADLs, and 49 AADLs	MMSE: 1) orientation to time and space; 2) memory registration; 3) attention and calculation; 4) recall; 5) language and constructive praxis.CAMCOG: orientation, language, memory, praxis, attention, abstract thinking, perception, and calculation.AADLs: IADL-x scale (9 items: 5 AADLS + 4 IADLs).Task 1: making toast and coffee; task 2: wrapping a gift as a present; task 3: preparing and packing a child's lunchbox and packing a child's schoolbag. i-ADL: household activities (9 items). a-ADL: evaluates complex activities.AADLs: 49 items Examples: crafts, use of complex technologies, and self-taught activities.	Greater executive dysfunction was associated with greater limitations in BADLs, IADLs, and AADLs. However, AADLs did not show stronger correlations than IADLs.The TMT-A is a significant indicator of BADL, IADL, and AADL performance.The CDT and AFT seem to contribute significantly to IADL and AADL performance.	Executive functions are less strongly related with BADLs than with IADLs and AADLs.The authors recommend using the TMT-A, CDT, and AFT as tool for indicating the need for a more detailed analysis of ADLs.	ADLs were assessed using a report-based method.There may have been a measurement bias in the reporting of ADLs (data were self-reported for controls, but reported by proxies for patients with cognitive impairment).Relatively small sample size.Measurements of executive functions were relatively simplistic.
Kalligerou et al.^45 (2020)Greece	1864	1551 HC223 MCI90 dementia	Study included in the HELIAD project	> 64 yearsHC: 72.4 years (6)MCI: 74.7 years (5.5)Dementia: 78.6 years (5.9)	To establish whether the inclusion of AADLs may increase the sensitivity of functional measurements to identify cognitive changes that may precede impairment of ADLs.	MCI (Petersen criteria)Dementia (DSM-IV-TR)AD (NINCDS-ADRDA)	ANOVAChi-square testLinear regression analysisCohort studyRetrospective, cross-sectional design	MMSEMCGCFT (non-verbal memory/visuoperceptual ability)Greek Verbal Learning TestSemantic and phonological verbal fluencySubtests from the Boston Diagnostic Aphasia Examination-short form (Greek-language version)Judgement of Line Orientation (abbreviated form)CDTTMT-ATMT-BVerbal fluency, Anomalous Sentence Repetition, Graphical Sequence Test, Motor Programming (Lezak).IADL-x scale (9 items)IADL scale (7 items extracted from the PGC-IADL and BDRS)	Cognitive functions: memory, language, attention-speed, executive functioning, and visuospatial perception.AADLs: IADL-x scale (9 items: 5 AADLS + 4 IADLs). Examples of the AADLs assessed: visiting friends and family, voluntary work, going to a café or to a care centre to participate activities, going to the cinema or a restaurant or participating in sports, attending a class.	Patients with dementia presented greater difficulties than those with MCI and HCs, both on the IADL-x scale and in IADLs.In the analysis of HCs, lower IADL-x score was associated with poorer cognitive performance; this association was not observed with the original IADL scale.	Incorporation of more advanced IADLs in functional scales may be useful for detecting cognitive differences on the normal spectrum.	The results are based on cross-sectional data, which only establish the correlation between IADL-x scale performance and cognitive function in HCs, and not the tool's use as a predictor of future cognitive impairment.Dichotomous format of both functional scales.Longitudinal data from the study cohort are needed to demonstrate the value of the IADL-x as a predictor of cognitive impairment.
De Vriendt et al.^8 (2021)Belgium	130	47 HC39 MCI44 AD	GDH of UZ Brussels University Hospital	≥ 65 yearsHC: 77.94 years (5.26)(68–91)MCI: 82.08 years (5.24) (71–96)Dementia: 81.41 years (4.8/0) (74–92)	To demonstrate the functional decline in all 3 levels of ADL (BADL, IADL, and AADL); to test the hypothesis that this functional decline in ADLs may allow discrimination between HC, MCI, and dementia; and to evaluate the functioning of the BIA as a comprehensive tool encompassing all 3 levels of ADL.	HCs: patients at a GDH for non-cognitive complaintsMCI (International Working Group on Mild Cognitive Impairment criteria)AD (NINCDS-ADRDA)	One-way ANOVAChi-square testBonferroni testIndices and conditional inference treesLogistic regressionCase–controlCross-sectional	MMSECAMCOG-RBIA (6 BADLs, 9 IADLs, and 49 AADLs)	Cognitive functions: language (expression and comprehension), orientation, memory, attention, calculation, registration, praxis, and recall.AADLs: 49 items.Examples: crafts, use of complex technologies, and self-taught activities.	Discrimination between HC and MCI and between MCI and AD was moderately successful with IADLs, in addition to age.	Although the BIA cannot currently be used as an independent diagnostic tool for cognitive disorders, it plays an important role in research into all-encompassing, multidisciplinary diagnostic research that addresses medical, cognitive, psychological and functional aspects.	There may have been a measurement bias due to the use of self-reported ADL data in HCs and proxy-reported data in MCI and AD.The study used a small sample size, and is therefore sensitive to small changes and does not support robust conclusions.Another potential bias: HCs were “apparently” cognitively healthy, but some may have had mild cognitive issues.
González et al.^47 (2022)United States	18,097	37% men (n = 6754)63% women (n = 11,343)	38 Alzheimer disease research centres belonging to the NACC	Time 1: mean age = 71.24 years (10.17)Time 2: mean age = 72.50 years (10.12)	To establish specific measures of syndrome stage (from absence of cognitive impairment to severe dementia) for functional change (establishing multi-stage distributional and reliability statistics for a functional measure, and using these to create useful reliable change indices and clinically meaningful differences, stratified by syndrome stage).	Patients previously diagnosed at the participating centres (from lack of cognitive impairment to severe dementia).FAQCDR	Wilcoxon signed rank testCohen's dNon-parametric Spearman rho testIntraclass correlation coefficientJacobson and Truax formula and calculation of the model standard error of estimateROC curveCohort studyLongitudinal (2 visits) 2005 and 2020	CDRFAQ	CDR: memory, orientation, judgement and problem solving, social activities (AADL), home and hobbies (partial AADL), and personal care.FAQ: items 5,6,7, and 10 assess AADLs; items 9 and 11 assess both IADLs and AADLs, in different proportions.	Marked differences were observed in the distribution of functional ratings according to syndrome stage.Differences in functional change were also associated with the progression through syndrome stages.These results were used to develop stage-specific metrics for reliable change indices and clinically meaningful differences.	Indices provide a method that was not previously available, enabling physicians to determine whether the observed functional change is reliable or meaningful, according to syndrome stage.The study provides a straightforward metric for assessing clinically significant impairment, which may be used to inform the follow-up of the disease and planning of treatment. In this regard, despite the need for further work in this area, this study provides sensitive, specific markers of reliable and meaningful changes in measures of functional outcomes, which may improve longitudinal clinical staging (including the advance through stages from absence of cognitive impairment to MCI).	The NACC dataset is not drawn from an epidemiological study and is not representative of the general population of the United States.The sample is predominantly made up of older, white, English-speaking individuals with a high level of schooling, limiting the extrapolation of findings to other populations.Collateral sources may be co-present in the individual, with closer relationships (paid caregiver, spouse, adult child) or higher levels of education being associated with higher scores in the more frequent questions.The time interval between the 2 assessments was not uniform, and varied greatly for some individuals; furthermore, the NACC does not report detailed information to explain this variability; therefore, this specific characteristic of the sample is not known.The study does not analyse the differences in reliable change indices and clinically meaningful differences according to the suspected aetiology of cognitive impairment; therefore, it is possible that some subtypes of dementia may vary in terms of the progression of functional impairment between syndrome stages.

Abbreviations: AADL: advanced activities of daily living; a-ADL: De Vriendt AADL assessment scale; a-ADL-CDI: a-ADL Cognitive Disability Index; a-ADL-DI: a-ADL Disability Index; a-ADL-PDI: a-ADL Physical Disability Index; NACC: National Alzheimer’s Coordinating Center; a-BT: abbreviated Barcelona Test; AD: Alzheimer disease; ADL: activities of daily living; AFT: Action Fluency Test; ANCOVA: analysis of covariance; BADL: basic activities of daily living; b-ADL: Katz BADL assessment scale; BDRS: Blessed Dementia Rating Scale; BIA: Brussels Integrated Activities of Daily Living Inventory; CAMCOG: Cambridge Examination for Mental Disorders of the Elderly, Cognitive part-original version; CAMCOG-R: Cambridge Examination for Mental Disorders of the Elderly-Revised, Cognitive part-revised; CAMDEX-R: Cambridge Examination for Mental Disorders of the Elderly-revised; CDR: Clinical Dementia Rating Scale; CDT: Clock-Drawing Test; CT: computed tomography; DSM-IV: Diagnostic and Statistical Manual of Mental Disorders, fourth version; DSM-IV-TR: Diagnostic and Statistical Manual of Mental Disorders Fourth version, Text Revision; FAB: Frontal Assessment Battery; FAQ: Functional Activities Questionnaire; GDH: geriatric day hospital; GDS/Reisberg: Global Deterioration Scale; GDS-Yesavage: Geriatric Depression Scale; GERRI: Geriatric Evaluation by Relatives Rating Instrument; HC: healthy controls; HELIAD: Hellenic Longitudinal Investigation of Aging and Diet; IADL: instrumental activities of daily living; i-ADL: Lawton and Brody IADL assessment scale; IADL-x: Instrumental Activities of Daily Living-extended scale; ICF: International Classification of Functioning, Disability and Health; IDDD: Interview for Deterioration of Daily Life in Dementia; Katz Index: Katz Index of Independence in Activities of Daily Living; MCGCFT: Medical College of Georgia Complex Figure Test; MCI: mild cognitive impairment; MMSE: Mini–Mental State Examination; MRI: magnetic resonance imaging; NAT: Naturalistic Activities Test; NINCDS-ADRDA: National Institute of Neurological and Communicative Disorders/Alzheimer’s Disease and Related Disorders Association; NORMACODEM: Normalization of Cognitive and Functional Assessment Instruments for Dementia; NPI-Q: Neuropsychiatric Inventory Questionnaire; PGC-IADL: Philadelphia Geriatric Center-Instrumental Activities of Daily Living, modified gender specific version; RDRS-2: Rapid Disability Rating Scale-2; ROC: receiver operating characteristic; SD: standard deviation; TMT: Trail-Making Test; WHICAP: Washington Heights-Hamilton Heights-Inwood Community Aging Project.

Regarding the characteristics of the study sample, all studies included healthy controls (HCs) and patients with MCI, and the majority also included patients with mild Alzheimer’s disease (AD) (some even included patients with moderate AD). The combined subject sample of all studies was 23,211; just over one-third of participants were men, nearly two-thirds were women, and the sample size ranged from 61 to 18,097. Age ranged between 60 and 96 years in all studies, with the exception of the study by Peña-Casanova et al.,³⁸ in which the mean age was 64.27 years (but with a standard deviation [DE] of 10.14 years).

Eight studies selected participants according to the National Institute of Neurological and Communicative Disorders-Alzheimer's Disease and Related Disorders Association (NINCDS-ADRDA) criteria. More specifically, 4 studies^8,39,40,46 additionally used the International Working Group on Mild Cognitive Impairment criteria and the Diagnostic and Statistical Manual of Mental Disorders (DSM) IV criteria (it should be noted that one of the studies by De Vriendt et al.⁸ also used the Cambridge Cognitive Examination [CAMCOG] and Mini–Mental State Examination [MMSE]). Furthermore, both studies by Cornelis et al.^43,44 used the Petersen criteria for MCI in addition to the NINCDS-ADRDA criteria. Two studies used CDR score as a selection criterion,^38,47 although the study by Peña-Casanova et al.³⁸ also used the NINCDS-ADRDA criteria, and in the study by González et al.,⁴⁷ in which all participants were diagnosed prior to inclusion, the Functional Activities Questionnaire (FAQ) was also used as a selection criterion in addition to the CDR.

The study by Kalligerou et al.⁴⁵ also used the NINCDS-ADRDA criteria, as well as the DSM IV-TR and the Petersen MCI criteria (the latter are the only criteria used in the study by Fieo et al.⁴²). Finally, the study by Bidzan et al.⁴¹ used the criteria of the International Working Group on Mild Cognitive Impairment.

In all studies, part of the sample included a subgroup of HCs, with the exception of the studies by Bidzan et al.⁴¹ (in which all participants were diagnosed with MCI) and González et al.⁴⁷ (which does not clearly state whether a subgroup of HCs is included, as all participants were drawn from a database in which they had previously been diagnosed with different clinical stages of cognitive impairment).

Statistical analysis

Regarding the methods of statistical analysis used, the majority of studies employed various of the same statistical indices and tests: the chi-square test is used in 7 studies, ANOVA in 5, ROC curve analysis in 4, and the Pearson correlation coefficient and Bonferroni test in 3; all 11 studies also use more specific tests, but none is used in more than one study.

Psychometric instruments

Regarding the psychometric instruments used to assess clinical and cognitive status, on the one hand, and functional status, on the other, cognitive assessment employed the MMSE in the great majority of studies (n = 8), and the CAMCOG in more than half (n = 6); the Trail-Making Test (TMT; parts A and B) is used in 2 studies, and more specific tests (or subtests) are used in various studies (eg, the Boston Naming Test to assess aphasia, the Clock-Drawing Test to assess frontal lobe function, etc). The study by Fieo et al.⁴² does not specify the cognitive instruments used. In more than half of the studies (n = 6), functional assessment included a new instrument assessing ADLs (some instruments exclusively measured AADLs, whereas some others also included items measuring, to some extent, IADLs and even some BADLs), which they sought to validate, comparing results against cognitive assessment. In the remaining 5 studies, the objective was not to fully or partially validate a functional assessment instrument for AADL, but rather to demonstrate the predictive validity of ADLs to assist in the diagnosis of MCI, similarly to the way that cognitive assessment is used. In some of these 6 studies, the novel functional assessment instruments were used as part of the functional assessment at the beginning of the research project (alongside cognitive assessment and other clinical tests), with the exception of the study by De Vriendt et al.,⁸ in which the Advanced Activities of Daily Living psychometric instrument (a-ADL) was not included as part of diagnosis. However, regardless of whether these instruments were included in the initial diagnostic process, the objective of the studies was their psychometric validation through comparison against cognitive assessment instruments. Therefore, some studies analysed the psychometric properties of these novel functional assessment instruments in order to assess their discriminative validity, predictive power in the diagnosis of MCI (and other disorders), and their convergent and concurrent validity with a view to establish their value in predicting the progression of a patient with MCI, for example. Some of these 6 studies also included other functional assessment instruments (in the initial functional assessment in each study), which already existed and are largely well-known, for example the Philadelphia Geriatric Center-Instrumental Activities of Daily Living (in 4/11 studies), the Katz Index of Independence in Activities of Daily Living (3/11) and other, more specific instruments assessing BADLs, IADLs, and AADLs (eg, the FAQ, Blessed Dementia Rating Scale [BDRS], Geriatric Evaluation by Relative's Rating Instrument [GERRI], etc) to complement and complete the comprehensive functional assessment of participants in the different samples. It should be noted that only the 2013 study by De Vriendt et al.⁸ additionally included psychometric instruments for psychopathological evaluation (Yesavage's Geriatric Depression Scale and the Neuropsychiatric Inventory Questionnaire) in the baseline assessment protocol, and only the study by Peña-Casanova et al.³⁸ used instruments (such as the BDRS) that include items screening for behavioural and personality changes; no other study used any psychopathological assessment instrument. Furthermore, only one study, the one mentioned above by De Vriendt et al.,⁸ included certain medical tests within the clinical, cognitive, and functional assessment protocol; these tests included physical examination, inventory of comorbidities and medicines used, blood analysis, and neuroimaging (this type of information is not reported in the remaining studies).

Relationship between cognitive function and AADLs

Regarding the cognitive functions and AADLs assessed in the different studies, it should be noted that although all studies (with one exception⁴²) included cognitive assessments in their protocols for the baseline diagnostic evaluation, only 8 of the studies reviewed directly compare cognitive functions and AADLs, whereas the other 3 do not directly compare the 2 variables, although they do evaluate AADLs. It should also be noted that, although the study by Fieo et al.⁴² does not specify the cognitive functions assessed (nor the cognitive psychometric instruments used), we are unable to affirm or rule out whether these were taken into account; furthermore, the other 2 studies that do not directly compare AADLs with cognitive functions (although they do include them in the baseline diagnostic protocol) were both performed by De Vriendt's working group,^39,40 which uses the a-ADL scale, presented for validation in a previous study by the same group.⁸

More specifically, all studies demonstrate to a greater or lesser extent that assessment of AADLs may be an important predictor of progression to AD, as observed in the study by De Vriendt et al.,⁸ for example. Approximately half of studies researched the efficacy of functional assessment of AADLs as a predictor of progression to MCI, among other outcomes, with some specific psychometric instrument created for the evaluation of AADLs, which is compared against gold-standard cognitive assessment instruments^{8,39,40,42,43,46}; in this regard, functional assessment instruments for AADLs (whether they evaluate AADLs only [eg, the a-ADL or Brussels Integrated Activities of Daily Living Inventory (BIA)] or include items evaluating AADLs [eg, the Instrumental Activities of Daily Living-extended (IADL-x)]) present optimal correlation with cognitive measures and moderate success in discriminating HCs from patients with MCI and between patients with MCI and those with AD.^{8,39,40,42,43,46} The a-ADL scale presented moderate to optimal efficacy overall, with significant discrimination between the HC, MCI, and AD groups (P < .01), and presented good psychometric properties (inter-rater reliability, agreement between patient and proxy, and correlations with cognitive tests); it was able to detect the subtle changes in functioning that occur in mild cognitive disorders, and to distinguish between the cognitive decline associated with normal ageing and that observed in MCI and AD.⁸ Within the a-ADL, the Cognitive Disability Index (a-ADL-CDI, which was strongly correlated with cognitive scales³⁹) and Disability Index (a-ADL-DI) subscales may be useful in the identification and follow-up of MCI in populations of older adults.⁸ Regarding the IADL-x, the authors observed that individuals scoring lower than the total median score in 6 activities from the scale presented 2.5 times greater risk of MCI at 2 years, and 2 times greater risk at 5 years; therefore, the instrument seems to be a valid tool for assessing the risk of developing MCI.⁴² Patients with dementia presented greater difficulties than those with MCI and HCs, both on the IADL-x scale and in IADLs. In analyses of cognitively healthy populations, lower IADL-x scores were associated with poorer cognitive performance; therefore, the use of functional assessment scales of AADLs may be helpful in detecting cognitive differences on the normal spectrum.⁴⁵

Other studies correlated items assessing AADLs (from more global or generic ADL assessment instruments) with cognitive assessment instruments or cognitive functions; in the same way that such psychometric instruments as the abbreviated Barcelona Test (a-BT) have been shown to be able to predict functional status,³⁸ functional assessment instruments of AADLs are able to discriminate normal age-related decline from that observed in MCI and AD, and are useful for detecting cognitive differences on the normal spectrum. Thus, the assessment of AADL function may represent an important predictor of progression to MCI and AD. Similarly, Bidzan et al.⁴¹ report that evaluation of the level of functional activity may be helpful in predicting the future progression of MCI.

Finally, some studies analyse the correlation between AADLs and cognitive functions with a view to establishing specific syndromic measures indicative of functional changes associated with clinical progression.^38,41,44,47 In this respect, Cornelis et al.⁴⁴ observed that AADL impairment increased in line with executive dysfunction, confirming the relationship between executive functions and AADLs, based on the results of cognitive instruments including the TMT-A, CDT, and Action Fluency Test (AFT). The authors found that these instruments were significant indicators of executive dysfunction and may significantly contribute to predicting a decrease in AADL performance, and argue that, as a result, there is a need for deeper analysis of ADLs, and specifically AADLs. Similarly, the study by Peña-Casanova et al.³⁸ observed a good functional correlation between a-BT global score and AADLs, at least among patients with MCI, reporting that all cognitive/functional correlations were statistically significant (P < .0001) and strong (ranging from 0.76 to 0.80), and concluding that the test would also enable prediction of functional status in this segment of patients. Finally, the study by González et al.⁴⁷ aimed to establish the differences in functional status associated with clinical progression through different syndromic stages using a (currently unavailable) method enabling physicians to determine whether the functional change observed was reliable or significant, according to syndromic stage, providing sensitive, specific indices of reliable and significant changes in measures of functional performance, which may help to improve longitudinal clinical staging (including the step from unimpaired cognition to MCI).

Discussion

Detailed analysis of the data reported by these studies revealed that all studies observed a more or less direct association between decline in AADL performance and the appearance of MCI, with many studies comparing the relationship between functional assessment of AADLs and cognitive assessment, with a view to establishing the predictive power of functional assessment in MCI diagnosis. Six studies reported that instruments assessing AADLs presented an excellent correlation with cognitive measures and good discrimination between HCs and patients with MCI (P < .01), and concluded that they generally present good psychometric properties. Other studies correlated items assessing AADLs with cognitive assessment instruments or cognitive functions, eg, associating executive functions with AADLs, and reported good validity, reliability, and sensitivity to change. This is demonstrated by the correlation between AADLs and cognitive function, and more specifically, executive functions (assessed using the TMT-A, TMT-B, CDT, AFT, and in many cases the a-BT). Furthermore, some studies report that functional outcomes may improve the clinical classification of MCI and the absence of cognitive impairment.

Other studies also related AADLs to cognitive functions, seeking to establish specific syndromic measures of functional changes in clinical progression, obtaining statistically significant results. Although the a-ADL, BIA, and IADL-x cannot be used as independent diagnostic tools for evaluating cognitive disorders compatible with potential MCI, they may be included in a comprehensive multidisciplinary assessment (including medical, cognitive, psychological, and functional aspects of the diagnosis of MCI), within the functional assessment of AADLs.

In the light of the above, and with respect to the objectives of the different studies included in the systematic review, all studies sought in one way or another to demonstrate the importance of considering impairment in AADLs (as highly complex functional activities, as regards their complex content, implicitly and explicitly involving cognitive, affective, functional, and social aspects) as predictors of progression to MCI,^{38,39,41,44–47} and the importance of including functional assessment of AADLs (through psychometric instruments and/or clinical interviews targeting this functional area) in clinical and neuropsychological assessment protocols for the evaluation of MCI, with contributions demonstrating the efficacy of this assessment in diagnosing MCI.^{8,38,40,42,43,46} Functional impairment of certain AADLs may be an early marker of cognitive impairment³⁹; consequently, follow-up evaluation of high-level functioning (AADLs) may constitute an important predictor of progression to AD.⁴⁰

Limitations

The main strengths of this systematic review lie in the inclusion and synthesis of information from diverse studies performed in different countries. This approach has resulted in an increased sample size, facilitating more reliable analysis of the results. Furthermore, the adoption of a systematic review methodology enables exhaustive exploration of the relevance of the idea that functional assessment of AADLs may be a sensitive measure for early detection of impairment of cognitive functions in the assessment of patients with possible MCI, and that, as a result, loss of AADLs may be a predictor of a process of MCI, alongside neurological and neuropsychological examination (ie, demonstrate the need to consider functional assessment of AADLs, at an equal level of importance to that of cognitive assessment in the evaluation of patients with possible MCI). In addition, the results of the systematic review demonstrate the need to monitor functional capacity in MCI (and the continuum from normal cognition to MCI) in order to ensure early diagnosis of MCI and to enable its prevention and treatment, which are of great importance in prevention programmes.

We should also consider the limitations of the study. Firstly, we identified a few studies (n = 11) addressing the subject in question; 3 of these^41,45,47 presented moderate (but acceptable) methodological quality. Furthermore, as noted in the methods section, we were not able to integrate the results quantitatively, in the form of a meta-analysis, due to the heterogeneity of study types and designs, study variables, sample sizes, types of patients, and outcome variables, etc.; therefore, we were only able to synthesise the scientific literature on the subject, hindering the integration of the results. Regarding study design, we included observational (cross-sectional and longitudinal) studies; no RCTs on the subject were identified. We also observed considerable variability in sample size, with 2 studies presenting relatively small samples^8,43 (although, given the objective of these 2 studies, psychometric validation of an instrument, the sample size selected may be considered statistically significant for that purpose) and others using larger samples. Furthermore, differences in the geographic settings in which the studies were conducted contribute to a certain variability, with cultural and contextual differences potentially influencing the interpretation of the results (eg, in the study by González et al.⁴⁷). Finally, differences in the year of publication of each study may have contributed to the heterogeneity of results due to societal changes over time, influenced by such factors as medical advances (for instance, the study by Peña-Casanova et al.³⁸ was the only study published before 2013).

There may also be a bias due to the heterogeneity of the clinical entity of MCI, in terms of its subtypes, and the variable reversibility/irreversibility, depending on the aetiology of the clinical syndrome.

Regarding the psychometric instruments used, it would be beneficial for studies to include evaluations at multiple timepoints: some studies, due to their cross-sectional design, present limitations in their ability to assess changes over time. However, within the framework of the creation of psychometric instruments, these cross-sectional studies do objectively demonstrate relationships between cognitive functions and AADLs, despite the fact that their cross-sectional design does not allow, in some cases, for confirmation of the hypothesis that AADLs may be a predictor of future MCI. More longitudinal studies are needed to address this issue.

Conclusions

While further studies (preferably longitudinal studies and RCTs) are needed in future, we should underscore the fact that the studies included in our systematic review support the idea that AADLs may be a sensitive measure for early detection of cognitive impairment. Furthermore, we may conclude that functional assessment of AADLs may detect cognitive differences between cognitively healthy individuals and patients with MCI, and may therefore be useful in predicting MCI and its future progression. Similarly, psychometric instruments for the functional assessment of AADLs may distinguish between normal age-related decline and the impairment observed in MCI and AD, and may be useful for detecting cognitive changes on the continuum from normal cognition to MCI (detecting the observable changes in functional status that occur in MCI) and are valid psychometric tools (as part of functional assessment) for assessing the risk of developing MCI.

Ethical considerations

This study complies with ethical standards. No human or animal experiments were performed, nor are any patient data presented; therefore, the study presents no risks to any person, and received all necessary authorisation for the use of different databases. Furthermore, the copyrights of all articles reviewed are respected in the literature review.

Funding

This study has received no funding from any public, private, or non-profit organisation.