Executive and General Cognitive Domain as a Relevant Factor of Specific Neuropsychiatric Symptoms in Alzheimer's Disease

ABSTRACT

Aim

To elucidate the pathophysiological mechanisms between neurocognition and neuropsychiatric sub‐symptoms in Alzheimer's disease (AD), the present cross‐sectional study compared severity of each subitem in a neuropsychiatric symptom (NPS) scale among four neurocognitive groups classified based on the pattern of executive and general cognitive function.

Methods

Of 546 consecutive outpatients who visited Memory Clinic at Jikei University Kashiwa Hospital, we selected 160 with AD and classified them into four neurocognitive groups based on score in the MMSE (Mini‐Mental State Examination: ≥ 21 point was general cognitive preserved) or FAB (Frontal Assessment Battery: ≥ 13 was executive cognitive preserved): BPC (both cognitive preserved); GCP (general cognitive preserved); ECP (executive cognitive preserved); NCP (neither cognitive preserved). We compared the severity of each subitem of the Behavioral Pathology in Alzheimer's Disease (Behave‐AD) scale among the four groups.

Results

Among seven subitems of Behave‐AD, the scores for diurnal rhythm disturbances and anxieties/phobias differed significantly among the four groups. The score for diurnal rhythm disturbances was significantly higher in GCP than BCP, and the score for anxiety/phobias was significantly higher in ECP than BCP. However, no significant difference was shown between each cognitive group and the NCP.

Conclusion

The general cognitive preservation without executive preservation may be relevant to the diurnal rhythm disturbances by self‐correction for own behavior, and the executive preservation without general cognition preservation may contribute to the transient anxiety as emotional reaction. Such dissociative relations between executive and general neurocognition may be relevant to specific neuropsychiatric sub‐symptoms emergence or severity in AD.

1. Introduction

As the most commonly encountered subtype of neurodegenerative disease, Alzheimer's disease (AD) is characterized by progressive impairment of neurocognitive functions (e.g., episodic memory disorder, impaired attention, and executive dysfunction), which result in progressive impairments of the activities of daily living (ADL) [1, 2, 3]. Apart from cognitive impairment, most patients with AD also exhibit neuropsychiatric symptoms (NPS) (e.g., psychosis, depression, anxiety, diurnal rhythm disturbances, and/or aberrant motor behaviors), which increase the distress or physical burden of the caregivers in the long‐term course of the disease [4]. According to previous studies, the pathophysiology underlying the emergence of the NPS involves interactions among psychosocial, demographic, and biological factors from the viewpoint of a suitable bio‐psycho‐social model [5, 6].

Among the causative factors mentioned above, from the neuropsychological aspects, previous studies have shown significant associations of neurocognitive impairments with the development of NPS [5, 7, 8, 9]. In particular, delusions, aberrant behaviors, and apathy, among the main neuropsychiatric symptoms have been reported as being related to the impairment of executive functions such as planning, performing, and self‐correcting, which implies that superior cognitive modalities are related to frontal‐subcortical circuits in AD [5, 7, 8]. On the other hand, general cognitive impairments such as impairment of memory, attention, and language, have also been shown as being associated with various neuropsychiatric symptoms, such as psychosis, aberrant behaviors, and apathy [9, 10]. Thus, arguments about the relevance of specific neurocognitive profiles in AD from cross‐sectional studies have been heterogeneous and diverse [9, 10, 11].

A systematic review conducted to examine the incidences of NPSs in the longitudinal course of AD showed that only the incidence of “apathy” among the neuropsychiatric sub‐symptoms consistently increased during the course of dementia, which implies that the correlative association between progressive apathy severity and neurocognitive deterioration (impairment grade) in the long‐term course [11]. Furthermore, more rapid cognitive declines in the course of AD were associated with the development of psychosis/aggression in the patients, implying that a poor compensative response to a faster neurocognitive slope may also be relevant to the emergence of confusional states, jumped conclusive ideation, and failure to correct own behaviors as neuropsychiatric sub‐symptoms in the longitudinal course of AD [12, 13, 14]. Taken together, it has been hypothesized that the gradient of the neurocognitive declines (directly) and the compensatory responses to such declines (indirectly) might be related to the development or worsening of neuropsychiatric sub‐symptoms and their diversity.

Of various neurocognitive impairments, executive function impairment in dementia has been the most frequently investigated in previous cross‐sectional studies as a critical neurocognitive domain relevant to the development of neuropsychiatric sub‐symptoms, and mono‐dimensional impairment of executive functions has been shown as being significantly correlated to the development of depression and anxiety [15, 16]. However, the severities of delusions, hallucinations, and disinhibition peaked in the mild or moderate neurocognitive stage in AD, which cannot be explained by presumption of a linear quantitative relationship between the neurocognitive stage and severity of NPSs [10]. Thus, we considered that examination of the association between the neurocognitive profile and development/severity each of the neuropsychiatric sub‐symptoms of AD might contribute to elucidation of the pathophysiological heterogeneity underlying the emergence of NPSs from multi‐dimensional neurocognitive viewpoints.

In the present cross‐sectional study, to elucidate the pathophysiological heterogeneity underlying the association between neuropsychiatric sub‐symptoms and the neurocognitive profiles in AD patients, we divided the study population into four neurocognitive groups based on the preservation or impairment of executive and general cognitive functions, and compared the severity/nature of neuropsychiatric sub‐symptoms among the four groups.

2. Methods

2.1. Participants Enrolled in the Present Study

The flow‐chart in Figure 1 depicts the procedure we adopted for selecting the sample for this study. Among the 13879 patients registered in our the database as new outpatients or inpatients at the Jikei University Kashiwa Hospital department of psychiatry between April 2006 and March 2014, we extracted a total 546 consecutive ambulatory outpatients who visited the Memory Clinic between August 2006 and December 2013, and selected 208 of these patients with AD (76.0% women, 78.1 ± 7.6 years old) [17]. The enrolled subjects were diagnosed as cases of probable AD based on the National Institute on Aging and the Alzheimer's Association (NIA‐AA) criteria [1]. All the diagnoses were made by a geriatric psychiatrist (one of the authors) based on a thorough medical history, physical and neurological examinations, routine blood tests, and findings of magnetic resonance imaging (MRI). The duration of illness (years) was defined as the term from the initial appearance of symptoms until consultation, based on information provided by reliable caregivers. Moreover, patients with insufficient data on the results of examination and/or the comprehensive medical history, a history of head trauma or brain damage, a history of drug or alcohol abuse, delirium, or who were receiving psychotropic medication, or delirium were excluded. The geriatric psychiatrist and clinical psychologist were experienced at performing neuropsychological and behavioral examinations, and validity in score of the scales was sustained by periodic discussions and exchanges of views. Finally, data of 160 subjects with AD were included in the statistical analysis. This study was conducted with the approval of the Ethics Committee of the Jikei University School of Medicine and Kashiwa Hospital.

FIGURE 1.

Flow diagram for sample selection from the database. ICD‐10, International Classification of Disease; NIA‐AA, National Institute on Aging and the Alzheimer's Association.

2.2. Assessments of the Neurocognitive Profiles and Neuropsychiatric Sub‐Symptoms

The Mini‐Mental State Examination (MMSE: score range, 0–30) was used to evaluate each patient's general cognitive functions, including memory, attention, language, and visuospatial cognition [18]. The Japanese FAB version, which consists of six subtests: (i) similarities; (ii) lexical fluency; (iii) motor series; (iv) conflicting instructions; (v) go/no‐go; and (vi) prehension behavior, was performed to evaluate the executive function in subjects [19]. Each subtest score ranges from 3 to 0, with the total score therefore ranging from 18 to 0 [19]. This neuropsychological test is easily administered at the bedside within 10–15 min by a clinical psychologist. To evaluate the severities of AD‐specific NPS based on a caregiver interview, we used the Behavioral Pathology in Alzheimer's Disease (Behave‐AD; score range: 0–75) scale [20]. The Behave‐AD consists of the following seven subscale domains: paranoid and delusional ideation (score range: 0–21); hallucinations (score range: 0–15); aberrant motor behaviors (score range: 0–9); aggressiveness (score range: 0–9); diurnal rhythm disturbances (score range: 0–3); affective disturbances (score range: 0–6); and anxieties and phobias (score range: 0–12) [20].

2.3. Assessments of the ADL Status

To examine individual basic or instrumental ADL (BADL or IADL) statuses, two scales have been utilized [21]. The Lawton IADL scale consists of five questions (ability to use telephone, shopping, mode of transportation, responsibility for own medications, and ability to handle finances) for men, and three additional questions for women (food preparation, housekeeping, and laundry). As the maximum IADL scores differ between women (score range: 0–8) and men (score range: 0–5) based on manual of original paper, we used the IADL percentage (%) in the present study, which was calculated as follows: IADL percentage (%) in men = (raw score)/5 points × 100 (%); IADL percentage (%) in women = (raw score)/8 points × 100 (%) [17, 21]. We also used IADL percentages to correct for sex differences in the domestic‐specific lifestyles in the present study. Moreover, to assess the BADL, we used the Physical Self‐Maintenance Scale (PSMS; score range: 0–6), which comprises the following six questions: toilet, feeding, dressing, grooming, physical ambulation, and bathing [21].

2.4. Classification Into Four Neurocognitive Groups Based on the FAB and MMSE Scores

To characterize neurocognitive subgroups from the two‐dimensional aspects of general cognition and executive functions, we classified the subjects with AD in this study (n = 160) into four neurocognitive groups based on the MMSE scores and FAB scores by reference to previous studies: the BCP group: both types of cognition preserved (MMSE score ≥ 21 and FAB score ≥ 13); the GCP group: only general cognition preserved (MMSE score ≥ 21 and FAB score ≤ 12); the ECP group: only executive functions preserved (MMSE score ≤ 20 and FAB score ≥ 13); and the NCP group; neither general cognition nor executive functions preserved: (MMSE score ≤ 20 and FAB score ≤ 12) [10, 22, 23, 24, 25, 26]. MMSE scores in the range of 21–30 points is conventionally regarded to correspond to mild AD: scores in a similar range have been used to classify the GCP group in clinical studies; FAB scores from 13 to 18 points have been used to define the ECP group according to a strict differentiation cutoff point between AD and amnestic mild cognitive impairment (aMCI) in previous studies as well as the present study [10, 22, 23, 24, 25, 26].

2.5. Statistical Analysis

To examine the statistical normality of each variable, the Shapiro–Wilk test was used, and the results showed that none of variables analyzed was normally distributed. Therefore, each of three demographic factors (age in years, duration of disease in years, and years of education), scores on two neuropsychological tests (MMSE and FAB), and two indicators of the ADL status (PSMS score or IADL percentage) as non‐parametric variables in the four neurocognitive groups were analyzed by the Kruskal‐Wallis test. Subsequently, the Mann–Whitney U test was performed as post hoc analysis to compare the variables between any two neurocognitive groups of AD and evaluate intragroup differences (pairwise comparisons). Sex difference (women or men) in the score was compared in the four groups by the χ ² test.

Next, the scores for each of the seven Behave‐AD subscales (paranoid and delusional ideation, hallucinations, aberrant motor behaviors, aggressiveness, diurnal rhythm disturbances, affective disturbances, anxieties and phobias) were compared and the total score as non‐parametric variables in the four groups were compared by the Kruskal‐Wallis test. Subsequently, the Mann–Whitney U test was performed as post hoc analysis to compare the variables between any two neurocognitive groups of AD.

In the multi‐group comparisons, p < 0.05 was considered as being indicative of a statistically significant difference. In the comparison of two groups as post hoc analysis, p < 0.05/6 (= 0.0083) was regarded as being indicative of statistical significance according to the Bonferroni correction. IBM SPSS Statistics for Windows, Version 22.0 (IBM Corp., Armonk, NY, USA) was used for all the statistical analyses.

3. Results

3.1. Characteristics of the Patients Classified Into the Four Neurocognitive Groups (Figure 2)

FIGURE 2.

Scatter plot between MMSE score and FAB score. MMSE, Mini Mental State Examination, FAB, Frontal Assessment Battery. Panel 1: Both cognitive preserved (BCP), Panel 2: General cognitive preserved (GCP), Panel 3: Executive function preserved (ECP), Panel 4: Neither cognitive function preserved (NCP).

Among the 208 patients with AD, data of a total of 160 subjects for whom MMSE and FAB scores were available were classified into four neurocognitive groups: (1) BCP group (n = 40); (2) GCP group (n = 24); (3) ECP group (n = 40); and (4) NCP group (n = 56), as shown in the scatter plot depicted in Figure 2. The demographic and neurocognitive variables in the 160 subjects were compared among the four groups (Table 1). Then, the total score and scores on the seven subscales of Behave‐AD were compared among the four groups. The mean (median) scores and standard deviations (SD) for the demographic characteristics are shown in Table 2.

TABLE 1.

Comparisons of demographics between 4 cognitive domains.

	(1) Both cognitive preserved (BCP)	(2) General cognitive preserved (GCP)	(3) Executive cognitive preserved (ECP)	(4) Neither cognitive preserved (NCP)	χ 2 score	p (4‐group comparison)	Post hoc comparisons (pairwise comparisons)
	(n = 40)	(n = 24)	(n = 40)	(n = 56)
	Mean ± SD range; Q1/median/Q3	Mean ± SD range; Q1/median/Q3	Mean ± SD range; Q1/median/Q3	Mean ± SD range; Q1/median/Q3
Age, years	79.2 ± 6.1 64–90; 77.0/79.0/83.0	77.4 ± 7.3 76–87; 74.3/78.0/82.8	76.9 ± 6.3 61–89; 72.5/78.0/81.0	77.7 ± 9.2 33–91; 75.0/78.0/84.5	3.1	0.378	(—)
Sex (women/men) a	12/28	7/17	9/31	13/43	0.9	0.817	(—)
Education, years	12.4 ± 2.4 6–16; 12.0/12.0/14.0	11.1 ± 2.2 7–16; 9.0/12.0/12.0	12.1 ± 2.5 6–18; 11.0/12.0/13.8	11.1 ± 2.8 6–16; 9.0/12.0/12.0	8.5	0.038*	(1) > (2)*, (1) > (4)*
Duration, years	3.1 ± 2.3 1–12; 1.0/2.0/5.0	2.7 ± 1.6 0–6; 1.3/2.0/4.0	3.0 ± 2.2 1–10;1.0/2.5/3.0	3.5 ± 2.5 0–10; 1.3/3.0/5.0	1.7	0.635
PSMS score (0–6)	5.8 ± 0.5 4–6; 6.0/6.0/6.0	5.6 ± 0.9 (6) 3–6; 6.0/6.0/6.0	5.2 ± 1.2 2–6; 5.0/6.0/6.0	4.9 ± 1.6 1–6; 4.0/6.0/6.0	8.9	0.031*	(1) > (3)*, (1) > (4)*
IADL score
Score range: women: 0–8/men: 0–5	6.0 ± 2.1/3.9 ± 1.1	5.5 ± 2.2/3.3 ± 1.0	5.2 ± 2.0/3.5 ± 1.8	4.6 ± 2.2/3.3 ± 1.9
IADL % (0–100)	77.7 ± 25.1 12.5–100; 60.0/87.5/100.0	69.2 ± 25.4 25–100; 39.4/75/90.6	67.3 ± 26.7 20–100; 45/75/87.5	59.7 ± 28.7 0–100; 37.5/60.0/87.5	7.0	0.071	(—)
MMSE score (0–30)	23.1 ± 1.9 21–27; 21.3/23.0/24.0	23.5 ± 1.7 21–27; 22.3/23.0/24.8	17.9 ± 2.1 13–20; 17.0/18.5.0/19.8	16.0 ± 3.8 4–20; 14.0/17.0/19.0	117.3	p < 0.001***	(1) > (3)*** ¶, (1) > (4)*** ¶, (2) > (3)*** ¶, (2) > (4)*** ¶, (3) > (4)*
MMSE‐Recall (0–3)	1.2 ± 1.2 0–3; 0.0/1.0/2.0	2.0 ± 1.2 0–3; 1.0/2.5/3.0	0.8 ± 1.0 0–3; 0.0/0.0/1.8	0.6 ± 0.9 0–3; 0.0/0.0/1.0	22.8	p < 0.001***	(2) > (3)*** ¶, (2) > (4)*** ¶ ¶, (1) < (2)*, (1) > (4)*
FAB score (0–18)	14.9 ± 1.2 13–17; 14.0/15.0/16.0	10.8 ± 1.3 8–12; 10.0/11.0/12.0	14.8 ± 1.3 13–18; 14.0/15.0/15.8	9.0 ± 3.3 0–12; 7.3/10.0/11.0	122.4	p < 0.001***	(1) > (2)*** ¶, (1) > (4)*** ¶, (2) < (3)*** ¶, (2) > (4)*, (3) > (4)*** ¶

Note: Kruskal–Wallis test as non‐parametric analysis was performed to compare all values between 4 cognitive domain groups. The Mann Whitney‐U test as a post hoc analysis was performed to compare between each cognitive domain. Values in bold font are significant results at each Bonferroni correction (p < 0.0083). *p < 0.05, **p < 0.01, ***p < 0.001, ^¶ p < 0.0083 = 0.05/6 (Bonferroni correction).

Abbreviations: FAB, Frontal Assessment Battery; IADL, Instrumental Activities of Daily Living; MMSE, Mini‐Mental State Examination; PSMS, Physical Self‑Maintenance Scale; Q, quartile; SD, standard deviation.

^a χ ² test was performed to compare the proportion of sex between 4 cognitive domain groups.

TABLE 2.

Comparisons of Behave‐AD sub‐scores between 4 cognitive domains.

Score range	(1) Both cognitive preserved (BCP)	(2) General cognitive preserved (GCP)	(3) Executive cognitive preserved (ECP)	(4) Neither cognitive preserved (NCP)	χ 2 score	p (4‐group comparison)	Post hoc comparisons (pairwise comparisons)
	(n = 40)	(n = 24)	(n = 40)	(n = 56)
	Mean ± SD range; Q1/median/Q3	Mean ± SD range; Q1/median/Q3	Mean ± SD range; Q1/median/Q3	Mean ± SD range; Q1/median/Q3
Total score (0–75)	4.7 ± 4.0 0–21; 1.0/4.0/7.0	6.7 ± 4.2 1–14; 3.0/5.5/10.5	6.6 ± 5.0 1–21; 3.0/5.0/9.8	6.5 ± 3.8 0–15; 3.0/5.0/9.0	7.3	0.063
Delusional ideation (0–21)	0.6 ± 1.5 0–8; 0.0/0.0/1.0	1.2 ± 1.8 0–6; 0.0/0.5/2.0	1.1 ± 1.8 0–7; 0.0/0.0/1.0	1.1 ± 1.7 0–6; 0.0/0.0/2.0	3.8	0.285
Hallucinations (0–12)	0.0 0.0–0.0	0.0 0.0–0.0	0.0 0.0–0.0	0.1 ± 0.6 0–4; 0.0/0.0/0.0	5.5	0.140
Aberrant motor behaviors (0–9)	0.9 ± 1.1 0–5; 0.0/0.0/1.0	1.0 ± 1.0 0–3; 0.0/1.0/2.0	1.0 ± 1.0 0–5; 0.0/1.0/1.0	1.2 ± 1.2 0–7; 0.0/1.0/2.0	3.5	0.316
Aggressiveness (0–9)	0.6 ± 1.1 0–5.; 0.0/0.0/1.0	1.1 ± 1.4 0–5; 0.0/0.5/2.0	0.9 ± 1.6 0–6; 0.0/0.0/1.0	1.1 ± 1.4 0–6; 0.0/1.0/2.0	3.9	0.275
Diurnal rhythm disturbances (0–3)	0.1 ± 0.3 0–1; 0.0/0.0/0.0	0.5 ± 0.7 0–3; 0.0/0.0/1.0	0.4 ± 0.6 0–2; 0.0/0.0/1.0	0.3 ± 0.6 0–3; 0.0/0.0/0.8	9.9	0.019*	(1) < (2)** ¶ , (1) < (3)*, (1) < (4)*
Affective disturbances (0–6)	0.8 ± 1.2 0–5; 0.0.0/0.0/1.0	0.6 ± 1.1 0–4; 0.0.0/0.0/1.0	1.0 ± 1.1 0–3; 0.0.0/0.0/2.0	0.6 ± 0.9 0–3; 0.0/0.0/1.0	3.4	0.334
Anxieties/phobias (0–12)	1.2 ± 0.9 0–4; 1.0/1.0/2.0	1.7 ± 1.2 0–5; 1.0/1.5/2.0	1.8 ± 1.1 0–4; 1.0/2.0/3.0	1.3 ± 1.4 0–8; 0.0.0/1.0/2.0	9.9	0.020*	(1) < (3)** ¶ , (3) > (4)*

Note: Kruskal–Wallis test as non‐parametric analysis was performed to compare the each Behave‐AD sub‐scores between 4 cognitive domain groups. The Mann Whitney‐U test as a post hoc analysis was performed to compare between each cognitive domain. Values in bold font are significant results at each Bonferroni correction (p < 0.0083). *p < 0.05, **p < 0.01, ^¶ p < 0.0083 = 0.05/6 (Bonferroni correction).

Abbreviations: Behave‐AD, Behavioral Pathology in Alzheimer's Disease; Q, quartile; SD, standard deviation.

3.2. Comparison of Scores MMSE and FAB in Four Neurocognitive Groups

The results of Kruskal‐Wallis test revealed significant differences of the MMSE and FAB scores among the four neurocognitive groups (p < 0.001). The scores in the two neuropsychological tests in the four groups were as follows: BCP group (MMSE: 23.1 ± 1.9; FAB: 14.9 ± 1.2); GCP group (MMSE: 23.5 ± 1.7; FAB: 10.8 ± 1.3); ECP group (MMSE: 17.9 ± 2.1; FAB: 14.8 ± 1.3); and NCP group (MMSE: 16.0 ± 3.8; FAB: 9.0 ± 3.3) (Table 1). Post hoc (Mann–Whitney U test) analysis of the MMSE scores showed that the score was significantly higher in the BCP group than that in the ECP group (p < 0.001) or NCP group (p < 0.001), significantly higher in the GCP group than that in the ECP group (p < 0.001) or NCP group (p < 0.001), and significantly higher in the ECP group than that in the NCP group (p = 0.026) (Table 1). Post hoc analysis of the FAB scores showed that the score was significantly higher in the BCP group than that in the GCP (p < 0.001) and NCP (p < 0.001) groups, significantly higher in the ECP group than in the GCP (p < 0.001) and NCP (p < 0.001) groups, and significantly higher in the GCP group than in the NCP group (p = 0.010) (Table 1).

Among the representative subitems of the MMSE scale reflecting memory function, the MMSE‐recall score was compared among the four neurocognitive groups, and significant differences were found (p < 0.001). The scores in the four groups were as follows: BCP group (1.2 ± 1.2); GCP group (2.0 ± 1.2); ECP group (0.8 ± 1.0); and NCP group (0.6 ± 0.9) (Table 1). Post hoc analysis of the MMSE‐recall scores showed that the score was significantly higher in the GCP group than that in the ECP (p < 0.001), NCP (p < 0.001), or BCP (p = 0.015) group, and significantly higher in the BCP group than that in the NCP group (p = 0.022) (Table 1).

3.3. Comparison of the Demographic Variables Among the Four Neurocognitive Groups

Among the demographic variables compared among the four neurocognitive groups, significant differences were found in the PSMS score (p = 0.031), with the results of post hoc analysis showing that the score in the BCP group (5.8 ± 0.5) was significantly higher than that in the ECP group (5.2 ± 1.2) (p = 0.014) and NCP group (4.9 ± 1.6) (p = 0.017) (Table 1). A significant difference was also found in the years of education (p = 0.038), with post hoc analysis showing that the years of education was significantly higher in the BCP group (12.4 ± 2.4) than that in the NCP group (11.1 ± 2.8) (p = 0.020) or GCP group (11.1 ± 2.2) (p = 0.036) (Table 1).

3.4. Comparison of the Total and Subscale Scores of Behave‐AD Among the Four Neurocognitive Groups

Significant differences in the scores for diurnal rhythm disturbances (p = 0.019) and anxieties/phobias (p = 0.020) were observed among the four groups as determined by the Kruskal‐Wallis test (Table 2) (Figure 3).

FIGURE 3.

Post hoc analysis (pairwise comparisons) of each subscale between two neurocognitive groups. Each score depicts mean (median) point. **p < 0.01, ^¶ p < 0.0083 = 0.05/6 (Bonferroni correction).

The scores for diurnal rhythm disturbances and anxieties/phobias in the four neurocognitive groups were as follows: BCP group (0.1 ± 0.3; 1.2 ± 0.9); GCP group (0.5 ± 0.7; 1.7 ± 1.2); ECP group (0.4 ± 0.6; 1.8 ± 1.1); and (4) NCP group (0.3 ± 0.6; 1.3 ± 1.4) (Table 2).

Post hoc analysis (Mann–Whitney U test) of the scores for diurnal rhythm disturbances revealed that the score in the BCP group was significantly lower than the scores in the GCP (p = 0.002), ECP (p = 0.011), and NCP (p = 0.029) groups (Table 2). The difference in the score for diurnal rhythm disturbances between the BCP and GCP groups was verified by the Bonferroni‐correction analysis as being robustly significant (Table 2) (Figure 3).

Post hoc analysis of the scores for anxieties/phobias score in the four groups showed that the score in the ECP group was significantly higher than that in the BCP (p = 0.007) and NCP (p = 0.011) groups (Table 2). Moreover, the difference in the score for anxieties/phobias was verified by score difference in group of BCP and ECP by the Bonferroni‐correction analysis as being robustly significant (Table 2) (Figure 3).

4. Discussion

In the present cross‐sectional study, we examined the association between the neurocognitive profile and the presence of neuropsychiatric sub‐symptoms in patients with AD. The results showed that among the seven Behave‐AD subscales, the scores for diurnal rhythm disturbances and anxieties/phobias differed significantly among the four neurocognitive groups (Table 2). Comparisons between any two groups followed by post hoc analysis revealed that the score for diurnal rhythm disturbances was significantly higher in the GCP group than the BCP group, and that the score for anxiety/phobias was significantly higher in the ECP than the BCP group. In spite of significant differences (p < 0.05) observed between any two among the four neurocognitive groups, Bonferroni‐correction analysis revealed no robustly significant difference between the NCP group and any of the other three neurocognitive groups (Table 2) (Figure 3). The FAB score in the BCP group was significantly higher than that in the GCP group (MMSE score preserved) and the MMSE score in the BCP group was significantly higher than that in the ECP group (FAB score preserved) (Table 1). Taken together, from the results described above, preservation of the general cognitive functions without preservation of executive functions may be associated with diurnal rhythm disturbances, and executive function preservation without preservation of general cognitive functions may contribute to the development of anxiety or phobias among the neuropsychiatric sub‐symptoms.

A previous study reported that the severity in the diurnal rhythm disturbances in cases of MCI with executive dysfunction was significantly greater than that in patients with any other type of MCI [15]. It is presumed that impairment of executive cognition in AD might be associated with poorer self‐correction of one's own distorted lifestyle, which could result secondarily in dysregulation of one's own behavior or circadian rhythm [27]. Inversely, sleep disturbances in AD were significantly associated with cognition and/or reduction in the ADL, and there may be a possibility of executive function decline caused by diurnal rhythm disturbances in AD [17, 28]. Therefore, the association between diurnal rhythm disturbances and neurocognition in AD may be correlative rather than causal [17, 28].

In some previous studies that investigated the relevance of neurocognitive decline in the development of neuropsychiatric sub‐symptoms, ‘executive dysfunction’ in subjects with MCI and AD was examined cross‐sectionally as the main neurocognitive domain other than memory function, and executive function decline was significantly correlated with the development of depression and anxiety [15, 16]. The poorer inhibition performance in patients with AD and executive function impairment in MCI were significantly associated with a greater severity of anxiety [15, 16]. However, the comparatively greater severity of anxiety in the initial stages of AD could be linked to preserved self‐awareness or insight into one's own general cognitive decline as an emotional reaction in the presence of preserved executive functions [29]. These results may be consistent with previous reports that memory impairment with preserved executive cognitive function is associated with depression, just as it is with anxiety, based on self‐underestimation of one's own neurocognitive impairments via hyper‐self‐awareness (hyper‐nosognosia) in the initial stage of AD [29, 30, 31].

The score for diurnal rhythm disturbances in the NCP group was weakly significantly higher than that in the BCP group, and the score for anxiety/phobias in the NCP group was weakly significantly lower than that in the ECP group (Table 2). Such results might suggest that a plateau point of severity of the corresponding neuropsychiatric sub‐symptoms is achieved in the GCP and ECP groups (Figure 3). Previous studies have shown similar peaking of the severity of neuropsychiatric sub‐symptoms (e.g., psychosis; anxiety; disinhibition) in the intermediate stage of AD, which implies that the sub‐symptoms might be temporary or transient reactive phenomena in the course of progression of neurocognitive decline [10, 25]. Thus, the present findings imply that progression of executive or general (e.g., memory and attentional function) cognitive impairments might result in shrinkage of reactive symptoms, including of diurnal rhythm disturbances, anxiety, and the like (Figure 3).

The present study had some limitations, such as the relatively small sample size. Previous studies have shown that executive function impairment in AD is significantly associated with the emergence or severity of psychosis and/or aberrant motor behaviors [7, 8]. In the present study, the scores for delusions and aberrant motor behaviors were higher in the GCP group than in the BCP group, although the differences were not statistically significant (Table 2). The present results may be in agreement with those of previous studies showing that impairment of executive cognitive function in AD is associated with the emergence/severity of psychosis and/or aberrant motor behaviors [7, 8]. On the other hand, the score for affective disturbances, including depression, was higher in the ECP group than that in the BCP group, although the difference was not significant (Table 2). These results might also lend support to a previous report that memory impairment with preserved executive cognition in AD is associated with depression, just as it is with anxiety [29, 30, 31]. To resolve the issue of a type‐2 error, a future study with a much larger sample size is needed. Second, the Behave‐AD was used to evaluate the AD‐specific NPS severity, but neuropsychiatric sub‐symptoms in this scale do not include “apathy” [20]. Apathy in AD has been shown in previous studies as being significantly correlated with neurocognitive declines [5, 10, 25]. Third, previous studies have reported the existence of significant association between the MMSE scores and FAB scores in the quantitatively correlation analysis, which means that it may be difficult to sustain qualitatively independent variables from two dimensionally neurocognitive viewpoins [32]. Fourth, comparisons of the demographic variables among the four neurocognitive groups revealed significant differences in the PSMS score (p = 0.031) and score for years of education (p = 0.038), but pairwise comparisons by post hoc analysis using Bonferroni's correction revealed no significant differences: p < 0.05/6 (= 0.0083) (Table 1). Therefore, the confounding contributions of PSMS score and years of education to neuropsychiatric sub‐symptoms were not investigated in the present study. Fifth, the Lawton IADL scale consists of five questions for men, and three additional questions (food preparation, housekeeping, and laundry) for women. The subjects enrolled in the present study were born predominantly before World War II (1945), which caused that traditional Japanese concepts may have distorted the reported distribution of housework and family roles. During the post‐World War II period and until the end of the high economic growth era (approximately the 1980s), the burden of housework was traditionally assigned to women—often referred to as housewives. Thus, the number of tasks assessed in the present study varied across genders, and there was a necessity to adopt a more representative scale (the Lawton IADL) that reflects the typical Japanese family structure post‐World War II in the present study. Finally, while it has been hypothesized that specific neurocognitive impairments are associated with the emergence of distinct neuropsychiatric sub‐symptoms, the reverse cause‐and‐effect, in which severe NPS can cause rapid cognitive declines remains unclear [13]. To resolve this issue, longitudinal research may be required, but neurocognitive impairments in AD could gradually progress or alter during the long‐term course, which may complicate these relationships. Therefore, we used only cross‐sectional data to clarify the relationships between neurocognitive decline and NPS.

Despite these limitations, the present study implies that the pathophysiological mechanisms underlying the emergence of specific neuropsychiatric symptoms in AD might involve the dissociative pattern of decline in executive and general cognitive functions. In the future, clarification of the mechanisms underlying the association between NPS and neurocognition may provide important information for selecting treatment stages that could lead to a reduction in the early caregiver burdens and prevent acceleration of the decline in ADL during the course of disease progression.

Author Contributions

Tomoyuki Nagata designed this study and have examined the subjects. Kiwamu Okabe wrote the manuscript. Shunichiro Shinagawa gave advice including the methods of statistical analysis, and reviewed the present manuscript. Kazutaka Nukariya have examined the patients with AD at the Kashiwa Hospital. Shinsuke Kito reviewed and commented on the final manuscript.

Funding

This work was supported by JSPS KAKENHI Grant Number 24K10690 and AMED under Grant Numbers JP24wm0625505.

Disclosure

The authors declare no conflicts of Interest.

Ethics Statement

This study was conducted with the approval of the Ethics Committee of the Jikei University School of Medicine and Kashiwa Hospital. No. 30‐308 (9329).

Okabe K., Nagata T., Nukariya K., Kito S., and Shinagawa S., “Executive and General Cognitive Domain as a Relevant Factor of Specific Neuropsychiatric Symptoms in Alzheimer's Disease,” Geriatrics & Gerontology International 26, no. 1 (2026): e70345, 10.1111/ggi.70345.

Data Availability Statement

The data that support 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.