Neuropsychological Correlates of Clinical Progression in Subjective Cognitive Decline

Abstract

Background:

Subjective cognitive decline (SCD) refers to self-reported cognitive decline in individuals with normal performance on standardized cognitive tests. Understanding the factors predicting progression from SCD to mild cognitive impairment (MCI) is crucial, as approximately 14% of SCD cases progress to dementia and about 27% develop MCI over four years.

Objective:

This study aims to identify neuropsychological predictors of progression from SCD to MCI, focusing on cognitive domains assessed through neuropsychological tests.

Methods:

This retrospective study at Seoul National University Bundang Hospital analyzed a cohort of 107 patients diagnosed with SCD through comprehensive assessment. Patients underwent annual neuropsychological testing, including the Digit Span Test, Boston Naming Test, Rey Complex Figure Test, Seoul Verbal Learning Test, and Stroop Test.

Results:

Annually, these patients underwent neuropsychological tests over a 5-year period; 24 progressed to MCI per NIA-AA criteria. Key predictors of MCI progression included age, ischemic heart disease, and scores from the forward digit span, delayed recall, and Boston naming tests. Lower scores in delayed recall and Boston naming tests significantly correlated with a higher risk of MCI (p < 0.001).

Conclusions:

These findings suggest a need for targeted management of memory and language functions to monitor disease progression effectively.

Keywords

Alzheimer’s disease mild cognitive impairment neuropsychological tests preclinical Alzheimer’s disease progressive subjective cognitive decline subjective cognitive decline

INTRODUCTION

Subjective cognitive decline (SCD) refers to self-reported cognitive decline in individuals that is not evident in objective cognitive assessments. According to the SCD-initiative (SCD-I) in 2014, SCD is characterized by a persistent cognitive decline experienced by the individual, unrelated to acute events, while still achieving normal performance on standardized cognitive tests used for diagnosing mild cognitive impairment (MCI), with adjustments made for age, gender, and education. 1

The prevalence of SCD in the elderly population is estimated to be around 50% to 60%, particularly among individuals aged 70 years and older. 2 While SCD was previously associated with affective symptoms or non-neurodegenerative causes, recent studies on the early stages of Alzheimer’s disease (AD) have gained credibility in linking SCD to the progression of AD.3 –6 Meta-analyses of longitudinal studies have reported that approximately 14% of individuals with SCD progress to dementia, and about 27% develop mild cognitive impairment over 4 years. 7

As there is currently no cure for AD, diagnosing SCD at an early stage and implementing proactive interventions are essential in slowing down cognitive decline. It is important to identify progressive SCD that may lead to MCI, as not all cases of SCD progress to MCI or demonstrate progressive AD pathology, as mentioned in previous studies.1,7–9 , 1,7–9 The objective of this study is to identify the cognitive domains associated with an increased risk of progression from SCD to MCI, using results from neuropsychological assessments. The SCD plus criteria enhance the standard SCD criteria by adding specific features that increase the likelihood of preclinical AD. This study aimed to investigate whether these criteria, characterized by subjective decline in memory rather than other cognitive domains, are evident in our cohort. In accordance with the criteria proposed by Jessen et al. 1 in their conceptual framework for research on SCD in preclinical AD, the SCD plus criteria were included to identify individuals more likely to have preclinical AD. However, this study also aimed to explore the clinical significance of cognitive assessments in identifying meaningful indicators in real-world clinical settings, where such biomarkers are not yet routinely available.

MATERIALS AND METHODS

This study analyzed a retrospective cohort of 107 patients who received an SCD diagnosis through a comprehensive assessment at Seoul National University Bundang Hospital. SCD was defined as self-reported cognitive decline with expressed concerns to the physician at the memory center, coupled with cognitive test performance superior to –1 standard deviation (SD) below age, sex, and education-adjusted norms across all subtests of the neuropsychological battery. Inclusion criteria for all groups also encompassed age ≥50 years and proficiency in the Korean language. The age criterion of 50 years was chosen to ensure a sufficient follow-up period for observing the progression of cognitive decline. Exclusion criteria included current depressive episode or medication, history of major psychiatric disorders, neurodegenerative disorders other than AD, vascular dementia, prior stroke with residual clinical symptoms, history of malignant disease, severe or unstable medical conditions, and clinically significant abnormalities in vitamin B12.

Diagnosis of SCD in all participants was conducted using the formal neuropsychological test battery SNSB, 10 comprising tests in five cognitive domains: the Digit Span Test 11 for attention, the Boston Naming Test 12 for language function, the Rey Complex Figure Test (RCFT) 13 for visuospatial function, the Seoul Verbal Learning Test (SVLT) 14 for verbal memory function, and the Stroop Test 15 (word reading and color reading) for frontal executive function. Scores in the 16th percentile or higher, corresponding to within –1 SD of the norm, were classified as normal. Follow-up assessments with the SNSB tests were conducted annually from the initial assessment.

The SCD plus criteria, 1 which include subjective decline in memory, onset of SCD within the last 5 years, age at onset ≥60 years, and concerns associated with SCD, were referenced for this study. However, APOE genotype testing was not routinely performed for the SCD group in this cohort, and thus, this biomarker was not included in our analysis.

The patients were monitored from April 2014 to September 2019. All participants were guaranteed annual follow-up with the SNSB. The determination of MCI progression was based on follow-up neuropsychological test results, following the clinical criteria for MCI due to AD outlined by NIA-AA. 16 The follow-up observation period for the progressive SCD group was defined as the time interval from the first neuropsychological test to the initial diagnosis of MCI, while the follow-up observation period for the stable SCD group was defined as the time interval between the first and most recent neuropsychological tests. Since SNSB tests were conducted annually, we extracted data from patients who underwent SNSB testing consistently every year over a 5-year period. Clinical information such as age, gender, education duration, presence of hypertension, hyperlipidemia, ischemic heart disease, and diabetes were anonymized and analyzed to understand the characteristics of the progressive SCD and stable SCD groups. This study was approved by the Institutional Review Board of Seoul National University Bundang Hospital (B-2106-692-103). The need for informed consent was waived because of the retrospective nature of the study.

For statistical analysis to examine the characteristics of each variable in the two groups, we conducted tests to assess normality for continuous variables such as age, education duration, and the results of each neuropsychological test. The Mann-Whitney test was performed for education duration, which did not follow a normal distribution. Independent t-tests were used to analyze other continuous variables. For binary variables such as gender, presence of hypertension, hyperlipidemia, ischemic heart disease, and diabetes, chi-square tests were conducted. Furthermore, to identify factors influencing disease progression, we employed Cox proportional hazard models to calculate hazard ratios. In order to minimize the effects between variables, multivariate analyses were conducted as follows. Model 1 utilized the Backward elimination variable selection method for all variables, while Model 2 included only variables that showed a significance level of 0.2 or lower in the univariate Cox regression analysis. For verbal learning test’s delayed recall and Boston naming test, which showed the highest correlation, binary variables were created based on the median values, and survival curves were plotted to observe the progression according to the high or low results.

RESULTS

A total of 107 patients received an SCD diagnosis and were followed up for 12–113 months (mean 52.93 months). Among them, six patients who did not undergo the complete initial neuropsychological test were excluded. Additionally, three patients with suspected conditions, such as cerebral ischemia, potentially impacting cognitive impairment, were excluded based on medical information verification. Furthermore, 30 individuals were excluded from the study because they did not complete the cognitive assessments or stopped attending follow-up visits. Finally, a total of 68 patients were included as the final study population. All patients met the criteria for SCD. 1 Among them, 24 patients progressed to MCI during the follow-up period and were classified as the progressive SCD group, while the remaining 44 patients were categorized as the stable SCD group. The 24 individuals who progressed to MCI from SCD accounted for approximately 35.3% of the sample.

Several variables demonstrated statistically significant correlations with progression to MCI, including age, ischemic heart disease, delayed recall, and recognition in the verbal learning test, and the Boston Naming Test (Table 1). The scores for each cognitive domain in the neuropsychological tests are reported as percentile scores, where a higher score indicates better function. Specifically, patients who progressed to MCI were approximately 4 years older on average compared to those who did not, and they had a higher prevalence of ischemic heart disease. Furthermore, significant score differences were observed in the verbal learning test, with approximately a 16-point difference in delayed recall and a 14-point difference in recognition. The MCI group also had a lower score of approximately 21 points in the Boston Naming Test. These findings suggest that among cognitive domains, memory and language ability were lower in the progressive SCD group.

Table 1

Basic characteristics of the study participants

	All	Stable SCD	Progressive SCD	p
	(n = 68)	(n = 44)	(n = 24)
Age, y	70.97 (51–86)	69.43 (51–83)	73.79 (64–86)	0.016
Female, n (%)	58 (85.3)	39 (88.6)	19 (79.2)	0.292
Education, y	11.39 (5–23)	12.083 (5–23)	11.011 (3–16)	0.353
Hypertension, n (%)	34 (50.0)	19 (43.2)	15 (62.5)	0.128
Diabetes mellitus, n (%)	12 (17.6)	6 (13.6)	6 (25.0)	0.240
Hyperlipidemia, n (%)	30 (44.1)	19 (43.2)	11 (45.8)	0.833
Ischemic heart disease, n (%)	8 (11.8)	2 (4.5)	6 (25.0)	0.012
Digit span test (% tile)
Forward	47.4551	51.2430	40.5108	0.144
	(2.28–99.20)	(13.57–95.95)	(20.29–74.74)
Backward	48.405	52.5567	58.9040	0.535
(3.75–99.82)	(39.08–80.98)	(34.46–99.82)
Verbal learning test, (% tile)
Delayed recall	52.9350	58.7816	42.2162	0.025
	(0.16–98.73)	(22.43–71.13)	(0.16–84.03)
Recognition	53.6764	60.0586	46.6713	0.030
	(2.57–92.34)	(3.36–92.34)	(2.57–91.90)
Boston naming test, (% tile)	50.0544	57.2982	36.7742	0.008
	(0.01–99.71)	(12.38–68.69)	(0.41–48.62)
Rey complex figure test, (% tile)	52.1019	51.7227	52.7971	0.876
	(0.00–89.44)	(11.50–84.49)	(31.96–83.71)
Stroop word test, n (%)^*	3 (4.6)	1 (2.4)	2 (8.7)	0.246
Stroop color test, n (%)^*	52.7742	56.9124	44.8741	0.117
	(0.94–94.41)	(0.94–82.91)	(17.19–90.21)

% tile, percentile. ^*Number and percentage of the subjects with <15 percentile score in the Stroop word test.

To identify the variables that increase the risk of progression from SCD to MCI, Cox regression analysis was conducted (Table 2). In Model 1, the following variables were consistently significant: age (HR = 3.520, 95% CI: 1.256–9.865), ischemic heart disease (HR = 30.939, 95% CI: 7.091–134.993), forward digit span test (HR = 5.737, 95% CI: 1.981–16.609), delayed recall from the verbal learning test (HR = 2.704, 95% CI: 1.014–7.215), and Boston Naming Test (HR = 8.103, 95% CI: 2.612–25.135). Among cognitive functions, the hazard ratio of the Boston Naming Test was notably high, suggesting that difficulty in language function is associated with a higher risk of progression to MCI in SCD patients.

Table 2

Association between factors and SCD progression

Variables	Univariate	Multivariate
	HR (95% CI)	Model 1	Model 2
		HR (95% CI)	HR (95% CI)
Age	1.913	3.520	–
	(0.823–4.449)^*	(1.256–9.865)
Sex, female	0.620 (0.029–1.839)	NI	–
Education^*	1.066 (0.419–2.711)	NI	–
Hypertension, yes	1.391 (0.600–3.227)	0.332	NI
		(0.107–1.031)^***
Diabetes mellitus, yes	0.882 (0.325–2.395)	NI	NI
Hyperlipidemia, yes	1.042 (0.456–2.381)	NI	NI
Ischemic heart disease, yes	8.982	30.939	0.050
	(3.056–26.397)^***	(7.091–134.993)	(0.011–0.241)^***
Digit span test
Forward	3.227	5.737	NI
	(1.322–7.7879)^*	(1.981–16.609)
Backward	0.927 (0.408–2.108)	NI	NI
Verbal learning test
Delayed recall	2.089	2.704	0.985
	(0.855–5.104)^*	(1.014–7.215)	(0.969–1.002)
Recognition	2.091	NI	0.966
	(0.885–4.939)		(0.979–1.014)
Boston naming test	3.808	8.103	0.966
	(1.499–9.676)^**	(2.612–25.135)	(0.946–0.988)^**
Rey complex figure test	0.930 (0.405–2.137)	NI	NI
Stroop word test	6.763	NI	NI
	(1.413–32.373)^*
Stroop color test	1.755	NI	NI
	(0.740–4.160)^*

Univariate and multivariate Cox regression analysis were performed. In Model I, a univariate analysis was conducted using all factors as input. In Model II, variables with a significance level below 0.200 from Table 1 were included in the analysis. In the analysis of Model II, age, sex, and education were included as adjusted variables. ^*p < 0.05, ^**p < 0.05, ^***p < 0.01.

In Model 2, controlling for age, sex, and education, significant associations were found for ischemic heart disease (HR = 0.050, 95% CI: 0.011–0.241) and the Boston Naming Test (HR = 0.966, 95% CI: 0.946–0.988). These results indicate that cognitive function, particularly language ability as measured by the Boston Naming Test, and cardiovascular health are critical factors in the progression from SCD to MCI.

To clarify, we constructed survival curves for variables that exhibited a high correlation with the risk of progression to MCI. Survival curves were generated for delayed recall from the verbal learning test and the Boston Naming Test, based on their respective median values (Supplementary Figure 1). The median percentile scores were 54.24 for delayed recall from the verbal learning test and 51.75 for the Boston Naming Test. These survival curves suggest that individuals with scores below the median in both cognitive domains are at a higher risk of progressing to MCI among SCD patients.

DISCUSSION

Over the 5 years of annual follow-up, we categorized participants into stable and progressive SCD groups, revealing that older age and lower scores in memory and language domains during neuropsychological assessments were associated with a higher risk of developing MCI. Notably, our research is the first to highlight the risk of SCD progressing to MCI based on low language ability using the Boston naming test, differentiating it from prior studies focusing on MCI to AD dementia progression. 17 Naming errors are recognized as an early indicator of AD, with evidence suggesting their presence even in comparisons between amnestic MCI patients and cognitively normal individuals.18,19, 18,19 The impairment observed in the Boston Naming Test supports the hypothesis of semantic breakdown due to storagedegradation. 20

However, our study encountered limitations. The Trail Making Test, previously mentioned as an indicating factor of progressive SCD, 21 was excluded in our study. Instead, we utilized the Stroop tests to assess executive function, but no statistically significant associations were found. Additionally, it is essential to note that the variable “ischemic heart disease” exhibited a remarkably high hazard ratio; however, due to the limited sample size, the statistical significance of this result could not be firmly established. Extensive research with a larger number of patients is necessary to address these limitations and gain a comprehensive understanding.

Another limitation is related to the exclusion criteria. Although current depressive episodes or medication were part of the exclusion criteria to avoid the inclusion of patients with pseudodementia, we did not statistically control for depression scores in our analyses. This could lead to selection biases, as individuals with a history of depression or those on antidepressant medications were excluded from the study. Consequently, we could not examine depression scores within the SCD group, potentially missing the impact of undiagnosed or denied depression on cognitive decline. Future research should include measures of depression to fully understand its impact on cognitive decline.

Additionally, we acknowledge the limitation related to variable follow-up periods and loss to follow-up. The variability in follow-up times among participants (12–113 months) could introduce biases. Some participants might have experienced cognitive decline but were lost to follow-up before this could be documented. Future studies should aim to minimize loss to follow-up and standardize follow-up periods to ensure more consistent data.

Despite these limitations, our findings have significant implications. We strongly recommend close monitoring of individuals with confirmed declines in memory and language, regardless of their cognitive function falling within normal neuropsychological ranges. Our findings can significantly aid clinicians in identifying potential causes and developing intervention strategies to improve daily functioning and overall well-being, especially in primary healthcare settings where invasive tests may not be readily accessible.

In conclusion, this longitudinal study emphasizes the critical role of language and memory impairment in predicting the progression from subjective cognitive decline to mild cognitive impairment. Careful neuropsychological assessment, with a focus on language and memory domains, can help identify individuals at high risk of cognitive decline.

AUTHOR CONTRIBUTIONS

Nayoung Ryoo (Writing – original draft); Jeewon Suh (Formal analysis; Investigation); Eun Ji Son (Conceptualization; Investigation); Young Ho Park (Methodology); SangYun Kim, M.D., Ph.D (Writing – review & editing).

Footnotes

ACKNOWLEDGMENTS

The authors express appreciation to the research participants and staff of the Memory Center of Seoul National University Bundang Hospital.

FUNDING

The authors have no funding to report.

CONFLICT OF INTEREST

The authors have no conflict of interest to report.

DATA AVAILABILITY

The data supporting the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.

The supplementary material is available in the electronic version of this article: .

References

Jessen

, Amariglio

, van Boxtel

, et al. A conceptual framework for research on subjective cognitive decline in preclinical Alzheimer’s disease. Alzheimers Dement 2014; 10: 844–852.

van Harten

, Mielke

, Swenson-Dravis

, et al. Subjective cognitive decline and risk of MCI: The Mayo Clinic Study of Aging. Neurology 2018; 91: e300–e312.

Hill

, Mogle

, Wion

, et al. Subjective cognitive impairment and affective symptoms: a systematic review. Gerontologist 2016; 56: e109–e127.

Mendonça

, Alves

, Bugalho

. From subjective cognitive complaints to dementia: who is at risk? A systematic review. Am J Alzheimers Dis Other Demen 2016; 31: 105–114.

Rabin

, Smart

and Amariglio

. Subjective cognitive decline in preclinical Alzheimer’s disease. Annu Rev Clin Psychol 2017; 13: 369–396.

Roehr

, Luck

, Pabst

, et al. Subjective cognitive decline is longitudinally associated with lower health-related quality of life. Int Psychogeriatr 2017; 29: 1939–1950.

Mitchell

, Beaumont

, Ferguson

, et al. Risk of dementia and mild cognitive impairment in older people with subjective memory complaints: meta-analysis. Acta Psychiatr Scand 2014; 130: 439–451.

Buckley

, Maruff

, Ames

, et al. Subjective memory decline predicts greater rates of clinical progression in preclinical Alzheimer’s disease. Alzheimers Dement 2016; 12: 796–804.

Caselli

, Chen

, Locke

, et al. Subjective cognitive decline: self and informant comparisons. Alzheimers Dement 2014; 10: 93–98.

10.

Ryu

and Yang

. The Seoul Neuropsychological Screening Battery (SNSB) for comprehensive neuropsychological assessment. Dement Neurocogn Disord 2023; 22: 1–15.

11.

Teng

, Wimer

, Roberts

, et al. Alzheimer’s dementia: performance on parallel forms of the dementia assessment battery. J Clin Exp Neuropsychol 1989; 11: 899–912.

12.

Kaplan

, Goodglass

and Weintraub

. The Boston Naming Test. 2nd Edition. Lea & Febiger, 1983.

13.

Tsatali

, Avdikou

, Gialaouzidis

, et al. The discriminant validity of Rey Complex Figure Test (RCFT) in subjective cognitive decline, mild cognitive impairment (multiple domain) and Alzheimer’s disease dementia (ADD; mild stage) in Greek older adults. Appl Neuropsychol Adult 2024; 31: 476–485.

14.

Kang

, Jahng

and Na

. Seoul Neuropsychological Screening Battery, 2nd Edition (SNSB-II). Seoul: Human Brain Research & Consulting Co, 2012.

15.

Vermeij

, Kessels

RPC

, Heskamp

, et al. Prefrontal activation may predict working-memory training gain in normal aging and mild cognitive impairment. Brain Imaging Behav 2017; 11: 141–154.

16.

Albert

, DeKosky

, Dickson

, et al. The diagnosis of mild cognitive impairment due to Alzheimer’s disease: recommendations from the National Institute on Aging-Alzheimer’s Association workgroups on diagnostic guidelines for Alzheimer’s disease. Alzheimers Dement 2011; 7: 270–279.

17.

Pravatà

, Tavernier

, Parker

, et al. The neural correlates of anomia in the conversion from mild cognitive impairment to Alzheimer’s disease. Neuroradiology 2016; 58: 59–67.

18.

Adlam

, Bozeat

, Arnold

, et al. Semantic knowledge in mild cognitive impairment and mild Alzheimer’s disease. Cortex 2006; 42: 675–684.

19.

Lin

, Chen

, Lin

, et al. Confrontation naming errors in Alzheimer’s disease. Dement Geriatr Cogn Disord 2014; 37: 86–94.

20.

Hodges

, Salmon

and Butters

. Semantic memory impairment in Alzheimer’s disease: failure of access or degraded knowledge? Neuropsychologia 1992; 30: 301–314.

21.

Fonseca

, Ducksbury

, Rodda

, et al. Factors that predict cognitive decline in patients with subjective cognitive impairment. Int Psychogeriatr 2015; 27: 1671–1677.