Recent contributions to the field of subjective cognitive decline in aging: A literature review

Abstract

Subjective cognitive decline (SCD) is defined as self-experienced, persistent concerns of decline in cognitive capacity in the context of normal performance on objective cognitive measures. Although SCD was initially thought to represent the “worried well,” these concerns can be linked to subtle brain changes prior to changes in objective cognitive performance and, therefore, in some individuals, SCD may represent the early stages of an underlying neurodegenerative disease process (e.g., Alzheimer's disease). The field of SCD research has expanded rapidly over the years, and this review aims to provide an update on new advances in, and contributions to, the field of SCD in key areas and themes identified by researchers in this field as particularly important and impactful. First, we highlight recent studies examining sociodemographic and genetic risk factors for SCD, including explorations of SCD across racial and ethnic minoritized groups, and examinations of sex and gender considerations. Next, we review new findings on relationships between SCD and in vivo markers of pathophysiology, utilizing neuroimaging and biofluid data, as well as associations between SCD and objective cognitive tests and neuropsychiatric measures. Finally, we summarize recent work on interventions for SCD and areas of future growth in the field of SCD.

INTRODUCTION

Subjective cognitive decline (SCD) is defined as the self-experienced, persistent decline in cognitive capacity compared with a previous normal cognitive status and is unrelated to an acute event.^1,2 Standardized terminology and criteria for SCD were first published in 2014 by the Subjective Cognitive Decline Initiative (SCD-I) group and, since then, several groups have worked to refine these diagnostic criteria and provide diagnostic considerations for health care professionals.^1–3 Despite these definitions, the term SCD is often applied and measured differently across different studies. For example, many research groups will use the term SCD to refer to a preclinical phase on the Alzheimer's disease (AD) continuum, although others will utilize the term SCD as a diagnostic representation of any self-reported cognitive decline, even if such decline is unrelated to an underlying neurodegenerative process. In addition, several different terms are used inconsistently across studies to refer to various aspects of SCD and the construct of SCD itself, such as “subjective cognitive concerns,” “subjective cognitive complaints (SCC),” and “subjective cognitive impairment.”

One of the challenges regarding consistent assessment of SCD as a diagnostic group involves the assessment tools that exist to measure SCD and SCC, as assessments may cover a variety of cognitive domains and span different timeframes of measurement and comparison (e.g., “over the last year” vs “over the last 10 years”; compared to participant baseline vs peers). Jessen and colleagues recently proposed that SCD measures should also assess for details including: (1) whether specific worries are associated with the perceived cognitive decline; (2) when the onset of the perceived cognitive decline occurred; (3) whether there is an association with any physical or mental conditions; and (4) whether there is an association with the use of medication, alcohol, or other substances.² Standardized methods of SCD assessment that have been validated with respect to objective outcome measures (e.g., biomarker data, objective cognitive performance, diagnostic progression, and so on) are needed in this field of research, particularly in the context of the large number of SCD measures available and the many potential underlying causes of SCD. Other factors that increase the heterogeneity of SCD may include sociodemographic and cultural differences in the understanding and expression of SCD among individuals (e.g., racial, ethnic, and gender diversity), various biological and environmental risk factors, as well as neuropsychiatric and genetic factors.

RESEARCH IN CONTEXT

• Systematic Review: The members of the SCD Professional Interest Area of the Alzheimer's Association International Society to Advance Alzheimer's Treatment were polled to assist with identifying key themes of particular impact to the field of subjective cognitive decline (SCD) in 2021 and critical studies within these themes. The authors then conducted a search via online scholarly repositories (e.g., PubMed) to review articles published over the last few years encompassing “subjective cognitive decline (SCD),” “subjective cognitive impairment,” or “subjective cognitive concerns.” Studies were prioritized if they fit within the predefined themes. Studies were only reviewed if they were published between 2021 and early 2022.

• Interpretation: This review aims to summarize recent contributions to the literature on assessment, diagnosis, and prognosis of SCD, as well as to identify areas of future study that would benefit the field.

• Future Directions: This review highlights that, although research on SCD has continued to rapidly expand over the past decade, further work is needed to refine diagnostic criteria and assessment methods, with consideration for the impact of culture and racial/ethnic diversity on diagnosis, as well as identify appropriate interventions for individuals with SCD.

The overall goal of this review is to summarize novel research in the field of SCD and how this work continues or adds to prior findings across several different themes and subtopics. This review aims to discuss such underlying factors that impact overall reporting of SCD, as well as to provide a brief synopsis of recent literature relating SCD to objective outcome measures (e.g., biomarker data, objective cognitive performance) and how these findings align with prior research or how they have carried our knowledge of SCD forward. Finally, we discuss avenues of future research necessary for better increasing our understanding of SCD and its relationship to the AD continuum. To ensure clear interpretation and consistency of summarized studies, throughout this manuscript we will use “SCD” to refer to the diagnostic group of subjective cognitive decline (i.e., individuals who have concerns about their cognition), whereas “SCC” will be used to refer to reported cognitive concerns (i.e., the number of, rates of, or types of concerns reported by individuals with SCD).

METHOD OF REVIEW

The primary focus of the literature review was centered on articles published between 2021 and early 2022, encompassing the terms of “subjective cognitive decline,” “subjective cognitive impairment,” or “subjective cognitive concerns.” Some manuscripts published in earlier years were included for the sake of comparison to prior research or to provide relevant background information for specific themes or subtopics. All members of the SCD Professional Interest Area (PIA) of the Alzheimer's Association International Society to Advance Alzheimer's Treatment were polled via email in 2021 to identify important themes in current SCD research, as well as to nominate recent articles in the field of SCD that they found most impactful. The results of this poll were used to establish the themes and subtopics covered in this review, as well as to provide additional guidance as to the most important works to highlight in this review. Several of the pre-identified themes were condensed during preparation of the manuscript to improve clarity and cohesiveness. The authors additionally used scholarly search engines (e.g., the National Institutes of Health [NIH] National Library of Medicine National Center for Biotechnology Information's “PubMed;” https://pubmed.ncbi.nlm.nih.gov) to identify all articles in the specified time frame with the above search criteria.

Over 700 manuscripts that met the above search criteria were identified in 2021 alone. We subsequently prioritized the inclusion of articles in this review that fit within the themes pre-identified by the SCD PIA poll as particularly impactful for the field, as well as those that included SCD as a separate and distinct diagnostic group, or directly examined SCC as a primary outcome measure (i.e., not focusing on articles in which SCD was merely a data point along the continuous AD spectrum and clear results related to SCD were not discussed). Although of course this article does not contain a summarization of every report published that included SCD or SCC in the specified timeframe, we aim for this review to provide a snapshot of current important areas of research in the field of SCD, as identified by members of the field, and how recently published work replicates prior findings and/or moves the field forward.

COGNITION AND RISK FOR FUTURE COGNITIVE IMPAIRMENT IN SCD

Conflicting reports exist as to whether SCD serves as an early marker of objective cognitive decline, and thus proves to be an imperative avenue to further investigate. As has been demonstrated previously, recent cross-sectional studies have linked SCD to lower cognitive performance in various domains compared to older adults without SCD,^4,5 although results that have found no cognitive differences between individuals with and without SCD have also been reported.⁶

Recently, the focus has shifted to studying the longitudinal relationships between trajectories of SCD and future cognitive decline/impairment. One such study did not detect any cognitive changes over 6 years in 40 individuals with SCD,⁷ although the use of a global cognitive screening tool as the primary outcome measure may have lacked the sensitivity to detect subtle cognitive changes that occur before the dementia stage. Sabatini and colleagues found that higher scores on an SCD questionnaire (i.e., greater levels of SCC) predicted 2-year decline on learning and memory tasks.⁸ However, looking at changes in cognition retrospectively from current measurements of SCC appears to be less helpful in detecting future change in SCC levels;. Gustavson and colleagues report that objective prior cognitive change over 10 years accounts for little variance in SCC ratings in individuals who were still cognitively normal at the time of initial assessment.⁹

At present, there are still conflicting results on SCD and its association with retrospective cognition, future cognitive decline, and risk for dementia. This could be attributed to the use of different measures to operationalize both SCD (as a construct)/SCC (as a symptom) and “objective cognition,” but also may suggest that examining SCD/SCC alone may neither be sufficient nor sensitive to predict objective functioning/declining. More work is needed to unpack and to explore the complex associations with other factors, such as sex,¹⁰ neuropsychiatric symptoms^11–14 and AD biomarkers.¹⁵ Recent research on relationships between SCD and these potential contributing factors will be discussed in subsequent subtopics.

SOCIODEMOGRAPHIC FACTORS AND SCD Racial and ethnic diversity and SCD

Observational research studies and clinical trials have historically struggled to recruit ethnically and racially diverse samples,¹⁶ which has limited our ability to understand the patterns of risk for SCD and future cognitive impairment in these individuals. New evidence suggests that racial and ethnic minority groups are at greater risk for AD and related dementias.^17,18 Recent investigations into the patterns of SCD across various sociocultural groups have sought to better understand how SCD presents across these historically marginalized populations. Such efforts may provide further insight into the early identification of pathological changes in neurodegenerative disorders among vulnerable populations and may enable us to observe unique risk factors for SCD among diverse populations.

The availability of data from large scale public health surveys, which include questions related to cognitive performance and functional status has allowed for evaluation of SCD across various demographic groups.^19–22 Recent data examining differences in SCD across diverse groups have been mixed. One study using the Behavioral Risk Factor Surveillance System (BRFSS) data, an extensive telephone survey of United States adults, to compare patterns of SCD in non-Hispanic White, non-Hispanic Black, and Hispanic adults 45 years of age and older found that non-Hispanic Black and Hispanic participants with SCD were more likely to be younger, have lower educational attainment, have lack access to health care, live alone, have more chronic medical conditions, and have greater SCD-related functional limitations compared to non-Hispanic White individuals with SCD.¹⁹ It is concerning that rates of discussing SCD with a health care provider were lower in the minority groups in this study despite higher rates of SCD. Along these lines, a separate study using the same data set found that Native Hawaiian/Other Pacific Islander (NHOPI) individuals had higher rates of SCD and SCD-related functional difficulties compared to Asian and non-Hispanic White respondents.²⁰ Another study evaluating SCD across U.S. Latino subgroups not only showed a greater percentage of reported SCD among Latino individuals relative to non-Hispanic White respondents overall but also significant heterogeneity in rates of reported SCD among Latino subgroups, indicating potential differences in reporting rates of SCD across different cultural or national groups.²¹ Zlatar and colleagues also reported significant variation in demographic characteristics and risk factors by background within a heterogeneous Hispanic sample with SCD (e.g., age, education, cardiovascular risk factors, anxiety, and depression).²³ Along these lines, recent data further suggest that not only are there differences between SCD reporting rates in Latinos/as/x compared to non-Hispanic White respondents, but there may be different relationships between SCC and objective cognitive performance between these groups.²⁴ On the other hand, another group demonstrated that associations between self-reported SCD and lower life satisfaction were similar across the various racial and ethnic groups studied.²²

In summary, recent research suggests that minority groups today face greater SCD-related challenges (possibly influenced by external stressors, cultural factors, or health factors in these groups) which, concerningly, are seemingly not discussed as frequently with health care providers and which potentially could limit early intervention for future cognitive impairment. These new data signal the importance of recruitment in the community to continue to develop our understanding of the complexities of SCD, particularly with regard to individuals from racially and ethnically diverse backgrounds, and to determine the most effective early intervention strategies for individuals from such backgrounds with SCD. In addition, more research is needed to understand relationships between race, ethnicity, and SCD in other countries, as many of the studies reviewed utilized samples based in the United States or other high-income countries. Recent work has shown that individuals in low- and middle-income countries are more at risk for higher levels of SCC if they have other chronic, comorbid medical conditions, including those that may cause reversible cognitive impairment.^25,26

Sexual orientation and gender identity considerations in SCD

Recent work has also highlighted differences in reporting rates of SCD in individuals of different sexual orientations and gender identities. Sex is defined as the biological differences that females and males have (e.g., chromosome, gonads, hormones, and reproductive functions). Gender, on the other hand, refers to the psychosocial frameworks in which individuals live and which can influence their subjective sexual identity.²⁷ Data regarding sex and gender differences in SCD have been mixed. A study by Brown and Patterson recently explored the moderation of gendered group status in the association with self-reported SCD and life satisfaction and found no differences across male- and female-gendered individuals.²² On the other hand, another study demonstrated that women with SCD tended to have lower education, lower premorbid intelligence scores, and were younger compared to men with SCD.²⁸ Similarly, a separate group examining individuals with autosomal dominant AD showed that females who carry a genetic mutation for AD endorsed greater levels of SCC compared to male carriers.¹⁰ It is also important to further understand differences in SCD among sexual and gender minorities, as a recent study showed that the prevalence of SCD was higher in individuals who identified as lesbian, gay, bisexual, and/or transgender or gender non-binary.²⁹ Similar to data from studies examining racial and ethnic diversity in individuals with SCD, this study also found similar rates of discussing SCD as a condition with health care providers between heterosexual cisgender adults and sexual and gender minority adults despite higher reporting rates of SCD in the latter group, as well as greater SCD-related functional limitations overall.

Overall, little had previously been known about rates of SCD in individuals of different sexual orientations and gender identity; however, recent research suggests that, similar to what has been observed in individuals from racially and ethnically diverse backgrounds, individuals who identify as sexual and gender minorities are more likely to report SCD, although may have lower rates of discussing their cognitive concerns with health care providers. This bourgeoning area of study is, therefore, important to not only ensure individuals who identify as a sexual and/or gender minority are adequately represented in our ongoing and future observational studies and clinical trials (particularly to study long-term outcomes in these individuals), but also to ensure that the best opportunities for screening, care, and treatment are being offered for these individuals at greater risk for reporting SCD.

Education, cognitive reserve, and SCD

Education has a complex relationship with SCD that may be influenced by other sociocultural and socioeconomic factors, including socioeconomic status.³⁰ Although educational attainment was associated with a greater likelihood of self-reported SCD among older adults with diabetes,³¹ this association was not independent of additional covariates, including marital and health insurance status, physical health status, and lifestyle factors.³² In addition, one recent study suggested that higher education is associated with earlier onset of SCD in females, although higher verbal intelligence was associated with later onset of SCD.²⁸

Education may also influence SCD by contributing to higher cognitive reserve, a term that refers to the differential susceptibility of cognitive abilities to aging, pathology, or brain insult across individuals through the adaptability of cognitive processes.³³ Although null associations have been reported,³⁴ higher education levels and higher levels of cognitive reserve have also been associated with reduced levels of SCC and lower risk for developing AD dementia later in life,^35,36 supporting previous findings from other cohorts.^37,38 There is also evidence that greater education levels may support cognitive performance in individuals in SCD,^39–41 and that this consistent association is independent of age, sex, and neurodegeneration.⁴² Finally, recent work suggests that education may protect against the accumulation of AD pathology in individuals with SCD, as higher education was associated with a reduced prevalence of amyloid positivity across 20 separate cohort studies, after adjusting for age, study setting and apolipoprotein E (APOE) ε4 status.⁴³

In sum, recent advances in the SCD literature, consistent with and building off prior findings, have indicated that although educational attainment has been associated with greater reporting rates of SCD or earlier onset of SCD, it is also associated with better outcomes in individuals with SCD, including lower levels of SCC, better objective cognitive performance, and lower amyloid burden. As most of the reviewed studies were cross-sectional, longitudinal work will be needed to establish firmer evidence of these protective effects in SCD and to better understand what mechanisms are underlying these complex relationships between education levels, cognitive reserve, and SCD.

GENETIC RISK FACTORS FOR SCD

In addition to demographic factors, recent work has indicated that genetic factors may increase the risk of SCD. The APOE ε4 allele, most commonly associated with AD dementia via increased levels of cerebral amyloid beta (Aβ), has also continued to show associations with changes in cognition and pathophysiological levels of amyloid and tau proteins in individuals with SCD.⁴⁴ In addition to associations with increased cerebral amyloid burden, recent data suggest that APOE ε4 allele carrier status in individuals with SCD may also be linked to other health factors associated with cognition and brain health, such as mean arterial pressure, number of white matter hyperintensities,⁴⁶ and functional brain network changes.⁴⁷ In APOE ε4 allele carriers and non-carriers with SCD, mild cognitive impairment (MCI), AD dementia, and a combination of vascular dementia and AD dementia, a dose–response pattern was observed across all groups, including the unimpaired SCD group, such that objective cognitive performance was incrementally worse as the number of ε4 alleles increased.⁴⁴ Another group found that, although SCC were associated with cognitive decline regardless of APOE ε4 carrier status, spouse-appraised memory functioning had stronger associations with objective cognitive decline in individuals who were ε4 carriers than non-carriers.⁴⁸

APOE is not the only gene implicated as a risk factor for cognitive decline among older adults with SCD. A recent study by Bessi and colleagues exploring the roles of CLOCK T3111C and PER2 C111G, polymorphisms involved in the sleep–wake cycle, found that the two individuals with SCD who eventually progressed to AD dementia were both PER2 C111G carriers (of eight carriers with SCD in total), whereas none of the non-carriers with SCD (n = 33) were diagnosed with dementia at the study end point.⁴⁹ A recent study sought to replicate findings from a 2017 study that initially compared unimpaired presenilin-1 (PSEN1) E280A mutation carriers, associated with early-onset autosomal dominant Alzheimer's disease (ADAD), to non-carriers, and found that levels of self-reported SCC were greater in carriers than in non-carriers, and that higher levels of study partner-reported SCC in carriers were associated with older age and lower hippocampal volume.⁵⁰ The results of this later study examining a larger sample of the same cohort was unable to replicate the findings regarding increases in self-reported SCC in carriers compared to non-carriers.⁵¹

In sum, recent research has shown links between certain genetic factors (i.e., alleles or mutations of APOE, CLOCK, PER2, and PSEN1) and worse objective cognitive performance, clinical progression to AD dementia, and/or increased AD biomarker burden (i.e., amyloid and tau) in individuals with SCD. That said, additional follow-up studies and longitudinal studies are needed to examine long-term progression rates to dementia and incident cognitive decline in individuals with SCD who carry such genetic markers, as some follow-up studies were unable to replicate previous findings regarding the link between SCC and genetic factors over longer time periods.

FLUID BIOMARKERS IN SCD

The study of fluid biomarkers has promise to identify additional tests and measures for the detection of pathophysiological proteins, inflammatory markers, and other indications of cerebral dysfunction that are less invasive and costly than traditional neuroimaging measures. Given that SCD is a preclinical phase of cognitive impairment, fluid biomarkers represent the possibility of useful, non-invasive, and low-cost methods to determine whether SCC are linked to underlying pathological changes prior to overt changes in cognition on objective assessments. Ideally, fluid biomarkers would be used as accessible and early diagnostic methods to determine which individuals with SCD are more likely to progress to stages of MCI or dementia.

As in the past, cerebrospinal fluid (CSF) measures of amyloid and tau have continued to be studied as markers of central nervous system pathophysiology, even in preclinical stages such as SCD, with consistent associations between AD-like profiles of CSF markers and progression from SCD to cognitive impairment, poorer cognitive performance, and reduced white matter integrity.^52–54

Additional work is being performed to identify alternative CSF markers that may be sensitive to early pathophysiological changes. An initial study examining N-224 tau found elevated levels in individuals with AD dementia compared to those with SCD, although N-224 levels were similar between individuals with SCD and healthy controls.⁵⁵ One group demonstrated a relationship between SCD and CSF neurofilament light chain (NfL), an indirect measure of axonal damage, such that SCD individuals had greater NfL compared to controls without SCD.⁵⁶ This relationship was moderated by Aβ status and linked to structural brain changes, such that SCD individuals with higher levels of Aβ tended to have greater NfL and reduced hippocampal volume. A similar relationship was also seen in a longitudinal sample of 304 post-stroke patients with SCD, where CSF circulating NfL predicted subsequent objective cognitive decline.⁵⁷ Elevated levels of CSF ferritin, a protein associated with inflammatory changes, have also been shown to have a dose–response relationship with AD-like profiles of CSF Aβ and total tau (t-tau) (i.e., greater levels of ferritin as AD pathophysiology and cognitive impairment progresses) in a group of healthy controls, SCD, and MCI individuals.⁵⁸

In addition, there has been substantial interest in AD plasma biomarkers in relation to SCD, given their accessibility and ease of collection compared to CSF. Some success has been seen in blood plasma measurements of phosphorylated tau (p-tau); for example, the clinic-based BioFINDER study demonstrated that plasma p-tau217 predicted AD dementia within 4 years with greater accuracy than clinical prediction by memory clinic physicians in a mixed group of individuals with SCD and MCI.⁵⁹ In addition, this group reported even greater accuracy after adding in other factors such as cognitive performance and genetic risk factors (i.e., APOE ε4 allele carrier status) and they further replicated these results in an Alzheimer's Disease Neuroimaging Initiative (ADNI) sample using plasma p-tau181. A separate group also found that elevated plasma p-tau217 levels were seen in individuals with SCD who had elevated Aβ levels on positron emission tomography (PET) scans.⁶⁰

However, certain plasma and biofluid markers may be less sensitive at discriminating preclinical SCD individuals from healthy controls. Separate studies examining CSF and plasma inflammatory levels and neutrophil gelatinase-associated lipocalin levels across healthy controls and individuals with SCD, MCI, and AD dementia found that CSF measurements of these markers were consistently able to discriminate between controls, SCD ,and cognitively impaired individuals.^61,62 However, plasma inflammatory markers were only elevated in the AD dementia group and there were no differences seen in plasma neutrophil gelatinase-associated lipocalin levels across diagnostic groups. Similarly, another study showed a dose–response relationship across SCD, MCI, and AD dementia groups using CSF measurements of t-tau, although t-tau levels measured using tear fluid in these individuals were only able to differentiate between SCD and AD dementia groups and were not sufficiently sensitive to detect changes between SCD and MCI individuals.⁶³

In sum, literature on the utility of biofluid-based biomarkers in the detection of underlying pathophysiology associated with AD and related dementias in early, preclinical stages such as SCD appears to be mixed. CSF markers of Aβ and tau continue to be studied as sensitive markers of underlying pathophysiology, even in preclinical individuals with SCD, including new markers of tau fragments, which may assist with the delineation between different types of tauopathies. Certain blood plasma markers of hyper-phosphorylated tau have also shown promising results in SCD and MCI individuals, although other blood plasma markers (e.g., inflammatory markers and markers of kidney dysfunction) and tear fluid tau levels seem to be less sensitive in discriminating between early preclinical stages of the AD process. That said, plasma markers represent an innovative area of research and opportunity for future studies to identify plasma biomarkers correlated with underlying cerebral pathophysiology in order to acquire less-invasive and costly biomarkers that are more accessible overall than traditional lumbar punctures and PET scans.

NEUROIMAGING IN SCD

SCD has been and continues to be linked to both structural and functional neuroimaging correlates. Compared to healthy controls, individuals with SCD show differences in both gray and white matter structure,⁶⁴ including reduced hippocampal and basal forebrain volume and cortical thinning in frontal, temporal regions,^65–69 as well as both local⁷⁰ and widespread differences in white matter.^65,71,72 These differences are related to subtle cognitive deficits in SCD.^{46,68,69,73,74} Recent studies have also highlighted the heterogeneous relationship between brain structure and SCD, which may be influenced by factors such as the study setting (i.e., clinical vs community).^64,75 The disease stage may also be a key moderator; at early stages, greater levels of SCC were associated with lower gray matter volume and reduced glucose metabolism, whereas at later stages of the AD continuum when cognition is impaired and insight is lessened, lower levels of SCC were related to greater neurodegeneration.⁷⁵

Functional abnormalities, as evident in task-free network connectivity and task-based neural activation studies, are also present in individuals with SCD. Higher SCC have been linked to decreased functional connectivity in the default mode network, as well as in the medial temporal and insular regions.^76–81 Recent findings have expanded the research and reported decreased functional connectivity in spatial navigation networks that included the right retrosplenial and right prefrontal cortices and the right retrosplenial cortex and right hippocampus,⁸² cognitive reserve networks defined by global connectivity to left frontal cortex,⁸³ and hippocampal networks in regions involved in olfactory function.⁸⁴ Moreover, an increasing number of studies in the field of SCD have begun to explore changes in dynamic functional connectivity,⁸⁵ which better captures the temporal properties of the functional network organization. The results have been somewhat inconsistent, with some studies reporting different dynamic brain patterns in individuals with SCD versus normal controls,^66,86 another study finding no differences,⁸⁷ a third study demonstrating more stable networks in SCD,⁸⁸ and finally another study finding regional differences in network variability.⁸⁹

Functional activation changes in SCD have also been linked to differences in cognition and physical functioning. When completing a memory task, people with greater SCC showed reduced hippocampal activity, even after adjusting for actual task performance,⁹⁰ and higher activation in the parahippocampal areas was found to predict better memory performance in individuals with SCD.⁹¹ Individuals with SCD appear to perceive memory deficits related to subtle neural changes that are subjectively sensible but cannot be detected by standard cognitive measurement,^75,90 validating that the basis of SCD is, at least partly, subtle objective cognitive decline.⁹² Meanwhile, cerebrovascular^93,94 and metabolic abnormalities^94–97 have also been noted in SCD and linked to increased functional signals observed in the frontal lobe in functional magnetic resonance imaging (fMRI) during N-back tasks.⁹⁴ Future studies may investigate the roles of different psychosocial and physiological factors that are related to increased and decreased functional activations in SCD.

On the AD continuum, SCD is characterized as the transition stage from normal to clinical AD. Accumulating evidence based on PET imaging studies confirms an association between higher Aβ and SCC.^90,98–101 Recent research efforts have explored the role of Aβ and tau in SCD-related neural changes^102,103 and found individuals with elevated Aβ and tau and SCD showed the greatest neurodegeneration. Taken together, these findings suggest that the presence of SCC in individuals with elevated Aβ may represent an early symptom related to incipient neurodegeneration.

There is also a rapidly growing and exciting literature on SCD involving the use of machine learning to best utilize multi-modal neuroimaging data in order to predict clinical diagnoses and disease progression in individuals with SCD.^104–109 In a large retrospective study, stroke history, lower education, lower score on the Montreal Cognitive Assessment cognitive screening measure, smaller left amygdala, and enlarged white matter at the banks of the right superior temporal sulcus were predictive of progression from SCD to MCI.¹¹⁰

Overall, although recent research on neuroimaging in individuals with SCD has continued to be consistent with prior findings, these new data have built off prior work to show the benefit of incorporating multi-modal imaging to detect relationships between SCD and sensitive measures of structural, functional, and pathophysiological changes in the brain. Leveraging multi-modal information including neuroimaging and clinical data has great potential to predict clinical progression and can benefit the field with more refinement and future research attention.

NEUROPSYCHIATRIC SYMPTOMS IN SCD

Neuropsychiatric symptoms (NPS) are commonly seen in dementia syndromes and can include depression, anxiety, apathy, agitation, delusions, hallucinations, and sleep disorders, among other symptoms.¹¹¹ However, as indicated by a recent clinic-based study that found that 81.4% of individuals with SCD had symptoms in at least one of these NPS domains per report (54% meeting the criteria for clinically relevant symptoms), NPS can also occur in preclinical stages of cognitive impairment.¹¹² Common NPS reported in individuals with SCD include: increased apathy, irritability, depression, anxiety, agitation, changes in subjective sleep quality, and mental rigidity.^112–118 Minor hallucinations in cognitively unimpaired patients with Parkinson's disease (PD) have also been linked to an increased likelihood of SCC and increased rate of brain atrophy within 2 years of diagnosis.¹¹⁹

However, NPS in SCD may not be entirely explained by early pathophysiological changes, as one community-based study estimated that 24.9% of the variability in reported SCC was explained by psychological and psychosocial factors.¹²⁰ Along these lines, elevated and worsening reports of SCC and NPS, including increased depression, anxiety, loneliness, social isolation, and general emotional distress, were reported during the coronavirus disease 2019 (COVID-19) pandemic despite stable objective cognitive performance.^121,122 Another study found that 34% of the variance in SCC reported during the COVID-19 pandemic was explained by a combination of perceived stress, negative emotions, and poorer general physical health related to prolonged confinement or quarantine.¹²³ Certain medical conditions may also be associated with elevated reports of SCC and increased NPS, including migraines and cerebrovascular disease, although the study examining cerebrovascular disease found that NPS were more associated with perceived SCD than cerebrovascular disease itself.^124,125

Data regarding the relationships between SCD, NPS, and cognition remain mixed. As stated previously, some studies found increases in SCC and NPS in the context of stable objective cognitive performance.^121,122 However, others have demonstrated that higher levels of anxiety, depression, and SCC are associated with lower performance on executive functioning and memory tasks.^8,11 In addition, one group showed that depressive symptoms partially mediated the relationship between SCD and objective memory performance, such that individuals with both depression and SCD were at risk for poor cognitive outcomes.¹³ Another group utilizing a sample of cognitively normal older adults only found a link between higher levels of SCC and worse objective cognitive performance when controlling for depressive symptoms.¹⁴ Differences in depressive symptom time of onset, duration, and severity among individuals across these studies may partially underlie these discordant findings.

The presence of SCD and NPS together has also been found to predict risk of future cognitive decline or dementia. A systematic review of studies examining NPS and cognitive decline demonstrated that individuals with SCD and anxiety, but not depression, were more at risk for future cognitive impairment.¹²⁶ Another group showed that comorbid depression and SCD had an additive effect on future risk of progressing to dementia, even when considering independent risks of these factors.¹²⁷ Over the past few years, research has also focused on a new diagnostic construct, mild behavioral impairment, to encompass functionally impairing, late-in-life-onset NPS across one or more domains of motivation, affect, impulse control, social appropriateness, and perception/thought content.¹²⁸ A recent study found that individuals with comorbid SCD and mild behavioral impairment were at greatest risk of incident cognitive decline over a 3-year period, compared to controls and those with either SCD or mild behavioral impairment alone.¹²⁹

Regarding relationships between structural brain changes, NPS, and SCD, findings have been mixed and may be influenced by severity of SCC. For example, NPS have been associated with medial temporal lobe atrophy in a mixed cohort of both SCD and MCI individuals.¹³⁰ These relationships may also be impacted by cerebral amyloid burden, as shown by two studies that showed correlations between greater worry or anxiety related to SCD and higher cerebral amyloid levels and, in one study, lower gray matter volume.^100,102 However, NPS and SCD may also be linked independently of structural and pathophysiological brain changes, as demonstrated by a study that found no alterations in hippocampal subfield volumes across groups of healthy controls and individuals with SCD and with primary depressive disorders, compared to reduced hippocampal subfield volumes in AD.¹³¹ Along these lines, a separate group demonstrated that elevated SCC levels were correlated consistently with higher levels of trait anxiety in a sample of 70-year-old individuals, and a cross-sectional relationship between higher SCC levels and Aβ positivity emerged only once anxiety was included in the model.¹⁰¹ It is worth noting recent work that demonstrated differences in relationships between SCD, NPS, and levels of Aβ across clinical and research samples and may help to explain these mixed findings.⁴³ In addition, study-partner reports of cognitive decline may be helpful in delineating cognitive changes related to NPS and cognitive changes related to underlying pathophysiology. Indeed, one study demonstrated that longitudinal study-partner cognitive complaints were more closely associated with in vivo AD biomarker levels that were participant complaints and were also less vulnerable to participant-reported NPS.¹³²

Overall, NPS may be present in preclinical stages of AD, such as SCD, and have been linked to structural and pathophysiological changes in the brain. However, the relationship between neurodegenerative disease and NPS is complex and difficult to disentangle, as data have shown that SCD and NPS can exist independently of any brain changes linked to neurodegeneration. Furthermore, the nature of the NPS themselves (e.g., depression vs anxiety) may differentially impact, mediate, and/or moderate relationships between SCD and structural, cognitive, and functional changes. Additional work is needed to improve our ability to discriminate between SCD associated with primary psychiatric syndromes in the absence of a progressive neurodegenerative process versus SCD and NPS co-occurring as early behavioral symptoms of neurodegeneration.

INTERVENTIONS IN SCD

Because the stage of SCD is increasingly recognized and understood to be a risk factor for cognitive decline, more interventions have been targeted to individuals in this preclinical stage prior to the onset of objective cognitive impairment. Given the paucity of treatment for neurodegenerative disorders, intervention at the point of SCD may be one of the earliest points of action with the potential of influencing trajectory and rate of progression. A recent systematic review of randomized control trials (RCTs) for interventions at the stage of SCD found that, overall, education programs, compared to cognitive training, physical training, or mind–body interventions, were most effective in improving memory function.¹³³ Several of these interventions, designed to support socialization and brain health education, appeared to support quality of life and holistic functioning in individuals with SCD.^134,135 Multimodal educational and health programs, involving combinations of brain health education, mindfulness practice, music listening, and/or exercise, were generally associated with increased subjective cognitive functioning post-intervention or reductions in anxiety levels.^136,137 Another recent systematic review also found strong evidence for cognitive/mental/physical training on cognitive and non-cognitive outcomes in SCD individuals.¹³⁸

There is mixed evidence supporting nutritional interventions in SCD, although stronger evidence for improved cognition and increased brain activity in individuals with SCD was seen in studies using shentai tea polyphenols, fermented dairy products (e.g., β-lactolin found in products such as yogurt), and combinations of nutrients, with less promising evidence for beneficial effects using other supplements such as vitamin D, specific proteins, and amino acids.^139–142

In addition, other studies examining the effects of exercise in individuals with SCD demonstrated improvement in fitness and mobility, but not necessarily improvement in subjective cognitive performance. In a study of older adults with SCD and comorbid hypertension, moderate and high intensity exercise programs improved fitness but not global cognitive function.¹⁴³ In a large four-arm RCT of exercise and exercise education programs in older individuals with SCD, participants in all programs showed improved gait parameters.¹⁴⁴

As various technologies have become ubiquitous, older adults are increasingly comfortable using a range of digital devices, opening the door to new treatment modalities that can result in improved cognitive functioning across various domains. Interventions for SCD using at-home digital devices and immersive virtual reality experiences have begun to be evaluated for feasibility and efficacy. Not only have these interventions been shown to be feasible, but evidence has shown that some computerized training programs can be associated with improved cognitive functioning in individuals with SCD.^145,146 Recent work has also demonstrated feasibility and enjoyability of an immersive virtual reality grocery store task to realistically measure general functioning in individuals with SCD.¹⁴⁷ In addition, an RCT examining biweekly virtual reality cognitive therapy reported positive increases in visuospatial function, apathy, affect, quality of life, and increased frontal-occipital functional connectivity in older people with SCD and MCI.¹⁴⁸

Other studies have examined the use of direct neuromodulation as an intervention for SCD. An RCT measuring the effect of a combined transcranial direct current stimulation (tDCS) and mindfulness-based stress reduction in individuals with SCD found that although feasible and safe, there were no statistically significant effects on outcome measures (i.e., mindfulness, social functioning, mood, subjective cognition and objective cognitive performance) compared to combined sham tDCS with mindfulness-based stress reduction.¹⁴⁹ A double-blind tDCS versus sham study found that individuals with SCD who had greater structural and functional brain integrity (per MRI) had the greatest tDCS-induced memory consolidation effect, indicating these individuals may benefit most from this form of neuromodulation.¹⁵⁰ A study of repeated transcranial magnetic stimulation in older adults with SCD found evidence of enhanced visual working memory, and significantly improved attention and executive function when the left dorsolateral prefrontal cortex was stimulated.¹⁵¹

In sum, data on recent interventions in SCD have been mixed. Educational programs have been highlighted as particularly effective for improving memory function in individuals with SCD, with good evidence for multimodal programs that also include psychological practices (e.g., mindfulness) and physical exercise as well as brain health education. Evidence has also shown that these programs can be delivered effectively via digital devices or computerized programs. On the other hand, data on specific nutrition/dietary supplements and transcranial stimulation for decreasing SCC and improving cognitive function have been more mixed and may depend on the specific supplement or intervention administered. Future research is needed to examine the long-term effects of these interventions on individuals with SCD, specifically as to whether they have any preventative or slowing effects on potential future cognitive decline and/or progression to dementia.

OPPORTUNITIES AND PRIORITIES FOR FUTURE RESEARCH IN SCD

Overall, there are numerous areas of opportunity for future research in the field of SCD, made evident through the work presented here and as highlighted by members of the Alzheimer's Association International Society to Advance Alzheimer's Research and Treatment SCD PIA. First, there has been a call for improved diagnostic processes for individuals with SCD for both clinical identification of individuals early in the disease process, as well as to serve as a potential outcome measure for clinical prevention trials. Several potential areas of improvement or future research in diagnostic processes have been identified, including the use of more universal and consistent assessments that capture cognitive decline across a variety of cognitive domains, more research on utilization of multimodal biomarker data in the diagnostic process (particularly for less invasive and costly biomarkers such as plasma tau measurements), and the utilization of both self- and study-partner reported data to accommodate diminished insight during the latter stages of AD. There are also other factors that need to be better understood to improve diagnostic accuracy of SCD, including SCD in cross-cultural settings, the impact of cognitive reserve factors such as education on SCD, and the relationships between NPSs and SCD. Finally, there is also a need to continue identifying appropriate interventions for individuals with SCD beyond psychopharmaceutical approaches, including cognitive, behavioral, physical, and neuromodulatory approaches, as well as a need for studies examining long-term effects of interventions on individuals with SCD (e.g., with data on future rates of progression to cognitive impairment/dementia or cognitive decline).

ACKNOWLEDGMENTS

There was no funding for this review article.

CONFLICT OF INTEREST STATEMENT

None of the authors (C.E.M, R.B., X.C., G.C., C.G., R.J.J., J.M., I.O., T.R., E.W., Y.T.Q., J.R.G., P.V., and R.A.) have any relevant disclosures or conflicts of interest. Author disclosures are available in the Supporting Information.

CONSENT STATEMENT

No human subjects were recruited for this review of previously published works, and, as such, additional consent was not necessary. Details of informed consent for each study included in this review can be found in the original publications.